SlideShare une entreprise Scribd logo
1  sur  5
Télécharger pour lire hors ligne
Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36
www.ijera.com 32 | P a g e
S
Text line Segmentation of Curved Document Images
Anusree.M*
, Dhanya.M.Dhanalakshmy**
*
(Department of Computer Science, Amrita Vishwa Vidhyapeetham, Coimbatore -641 112)
**
(Department of Computer Science, Amrita Vishwa Vidhyapeetham, Coimbatore -641 112)
ABSTRACT
Document image analysis has been widely used in historical and heritage studies, education and digital library.
Document image analytical techniques are mainly used for improving the human readability and the OCR
quality of the document. During the digitization, camera captured images contain warped document due
perspective and geometric distortions. The main difficulty is text line detection in the document. Many
algorithms had been proposed to address the problem of printed document text line detection, but they failed to
extract text lines in curved document. This paper describes a segmentation technique that detects the curled text
line in camera captured document images.
Keywords – Curved document images, Gradient vector flow, Gray scale image, Optical character
recognition, Rectification.
I. Introduction
Digitization of document is the most
commonly used method for increasing the quality and
compatibility of the document. Now a days,
compared to scanned document, camera captured
document has been widely used for digitization. Most
of the camera captured documents such as degraded
books and newspapers suffer curling of text lines due
to perspective distortion. There is a huge amount of
document images in libraries and in various national
archives that have not been exploited electronically.
If these documents need to be electronically
available, the first and major step will be document
segmentation into text lines. Before text line
segmentation, preprocessing steps are carried out.
Thresholding is used for the binarization that
converts a gray-scale (8 bit per pixel) document
image into a binary document image (1 bit per pixel).
It can be divided into global thresholding and local
thresholding. For flattening or straightening the
document, first stage is to segment the text line from
the document images. This is the most difficult step
in rectification method. There are many methods in
text line extraction in the typical text line extraction
stage. But in the case of the curled document image,
it is difficult to use usual algorithms such as
projection profile, x cut, run length smearing etc.
Here we are proposing method text line segmentation
for curved document images.
II. Literature survey
There are many methods to extract text
lines from single-oriented documents. Shijian Lu et
al. [1] proposed a method for implementing the text
line segmentation of curved document. The
rectification of curled text lines has been carried out
by using partition based approaches. Partition based
approaches include baseline detection, x line, and
vertical stroke identification. Here x line and baseline
have been estimated using least square fitting of the
character tip point. Documents are partitioned into
multiple quadrilateral patches. Finally target
rectangle has been constructed. It is done using
aspect ratio and number of the characters enclosed
within the partitioned block. This method work on
Latin-based document and the authors classified the
document in to six categories based on shape such as
character span, character ascender and descender.
Compared with the existing algorithms, this method
doesn’t need any special hardware or camera
calibration. But the main disadvantage of this system
is that it will not work on low resolution images. This
method is not suitable for handwritten and historical
documents.
Syed Saqib Bukhari et al. [2] presented a new
algorithm for text line segmentation of curled
document images. Here curled text-line segmentation
is performed by adapting active contour (snake)
which has been used for de-wrapping and improving
the OCR quality as well as the readability. Snakes are
used for estimating the local pair of x line and base
line. This techniques uses x line and base line
estimation using connected component analysis and
this is applied to the monocular single camera
captured document images. The overlapping x line
and baseline are considered as segmented text lines.
Here coupled snakes model has been used instead of
snake model. The snake moves towards targeted
object under the influence of internal energy and
external energy. The internal and external energy has
been calculated from image content and snake point
RESEARCH ARTICLE OPEN ACCESS
Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36
www.ijera.com 33 | P a g e
respectively. After finding the external energy, the
evolving and deformation of the snakes have to be
carried out using gradient vector flow. The main
advantage of this method is that it is less sensitive to
high degree of curl and skew, while the challenge is
that it produces many number of over-and under-
segmentation errors.
Partha Pratim Roy et al. [3] proposed text line
extraction using foreground and background
information in the document image. For the
extraction of curved document images, water
reservoir technique has been used. Using positional
information and the size, each individual component
is grouped together as a cluster. In the second stage,
the potential region has been found and individual
lines are estimated using candidate region. The main
advantage of this method is to determine the
document containing multi-oriented and curled lines.
The steps include initial character clustering,
grouping cluster, candidate point extraction followed
by extension of cluster group. The efficiency of the
algorithm depends upon the binarization result. So it
does not work with the colored document. But it is
independent of font style, rotation, scale etc..
Vaibhav Gavali et al. [4] describe the algorithm that
is independent of the text and the color of the
document. The paper describes a technique which
extracts text lines from a curved gray scale images.
The gray scale image has been taken as input;
binarized and discrete haar wavelet is applied
followed by thresholding to remove some non-text
region in the document. The morphological operation
has been used for each band to connect disconnected
characters. At last, 3 dilated bands are combined
using logical AND operator to get segmented text
lines. Here 2D haar wavelets are used for edge
detection which decomposes the input document in to
LL, LH, HL, and HH. The detailed information of the
text lines are presents in the LH, HL, HH and this
output is threshold using dynamic threshold.
G. Louloudis et al. [5] used a block-Based Hough
Transform for text line extraction. This method is
used for unconstrained handwritten documents. In the
first step, preprocessing for enhancement of image,
connected component extraction and average
character height estimation were carried out. In
binarization, adaptive thresholding method has been
used. From the result all connected component has
been identified and area of the bounding box is
calculated. The proposed approach consists of
partitioning the connected component space into
three subsets, and the splitting of connected
components into equally spaced blocks each of them
voting in the Hough domain. This proposed method
is applied to handwritten document which are in the
Latin format. The output of the document is post-
processed to reduce more than one line which
corresponds to same Hough domain.
III. Proposed method
The architecture of the system is shown in
the figure 1. In the first stage, input color image is
converted to gray scale and preprocessed. The next
stage determines edges. At last, curved text lines are
segmented using morphological operation.
Input image
Segmented text line
Fig.1.Proposed system architecture
3.1. Pre-processing
Pre-processing is the first step for document image
analysis. Here pre-processing include binarization
and noise reduction.
3.1.1. Binarization
In an OCR, one of the main processing stages is
binarization of document images. Binarization of a
text image should give the foreground text in black
and noisy background in white. There are different
thresholding methods already exist; they don’t give
exact results for all types of document images. Few
algorithms might work better for one type of
documents while they might give poor results for
other types where there is a complex background,
extremely low intensity variation. Depending upon
the document the output of threshold will vary, hence
in this project different methods were tried for getting
better result.
3.1.1.1. Otsu’s method
Otsu’s method is one of the famous global
thresholding methods. This method tries to find the
threshold T which separates the gray-level histogram
in an optimal way into two segments. Here the
calculation of inter-classes or intra-classes variances
is based on the normalized histogram of the image
H= [h0...h255] where Σ hi=1. In Otsu algorithm,
result has noise in the form of background being
detected as foreground.
3.1.1.2. Niblack’s algorithm
Niblack’s algorithm calculates a pixel-wise threshold
by sliding a rectangular window over the gray level
image. The computation of threshold is based on the
local mean m and the standard deviation s of all the
pixels in the window and is given by the equation 1
below
Preprocessing
Text line segmentation
Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36
www.ijera.com 34 | P a g e
skmT  )1(
Here m is the average value of the pixels pi, and k is
fixed to -0.2. Advantage of Niblack is that it always
identifies the text regions correctly as foreground but
on the other hand tends to produce a large amount of
binarization noise in non-text regions also.
3.1.1.3. Savola algorithm
Savola algorithm claims to improve Niblack’s
method by computing the threshold using the
dynamic range of image gray-value standard
deviation.The binarization is given by equation (2)
)
R
σ
k)(1m(1T  )2(
Where k is se to 0.5 and R varies from 0 to 255. This
method outperforms Niblack’s algorithm in images
where the text pixels have near 0 gray-values and the
background pixels have near 255 gray-values.
However, in images the gray values of text and non-
text pixels are close to each other results degrade
significantly.
Thresholding is used to segment an image by setting
all pixels whose intensity values are above a
threshold to a foreground value and all the remaining
pixels to a background value. The traditional
thresholding operator uses a global threshold for
entire images; this method changes the threshold
dynamically over the image. it can accommodate
changing lighting conditions in the image, such as
those occurring as a result of a strong illumination
gradient or shadows.
3.1.1.4. Adaptive thresholding
Adaptive thresholding takes a grayscale or color
image as input and it produces a binary image
representing the text lines in the image. The pixel
value is above threshold, it is set to the foreground
value; otherwise it assumes the background value.
Local adaptive thresholding will select an individual
threshold for each pixel based on the range of
intensity values in its local neighborhood pixels. This
allows for binarization of a document whose global
intensity histogram doesn't contain distinctive peaks.
The algorithm involves following steps:
1) Convolve the image with a suitable
statistical operator.
2) Subtract the original from the convolved
image.
3) Threshold the difference image with C.
4) Invert the threshold image.
Out of the four methods described above, adaptive
binarization algorithm was found to give better
results for our dataset.
3.1.2 Noise Removal
Morphological image processing is a set of
operations related to the shape or morphology of
features in an image. Here median filter is mainly
used for marginal noise reduction. The median filter
is mainly used for removal of salt and pepper noise. It
replaces value of the pixel by the median of the
intensity values in the neighborhood of that pixel.
3.2. Text line detection
Optical character recognition (OCR) or text
recognition is an important area in commercial
software development and academic research. The
application areas of digital document processing are
office and library automation, publishing houses,
communication technology, banking sector, historical
and many other domain areas. A large number of
techniques have been proposed to address the
problem of document text extraction. Text line
detection is a critical stage towards unconstrained
handwritten document recognition. It refers to the
segmentation of a document page image into distinct
entities, the text lines. The challenges in problems
that appear in this stage are the difference in the skew
angle between lines on the page, overlapping words
and adjacent lines touching.
3.2.1. Canny Edge Detection
To detect the edges in the document canny edge
detector has been used .The main advantages this
method are it has low error rate, edge point is
localized and accurate result is obtained on the single
edge point pixel. The canny detector algorithm
involves four steps:-
1. Smoothing the input image using guasssain
filter.
2. Find out gradient magnitude and angle
3. Nonmaxima suppreeion to the previous
result.
4. Apply double thresholding to identify the
edges.
The image smoothening is carried out by using
circular 2D Gaussian function which is denoted by
equation 3:-
2
22
2
)(
).(

yx
eyxG


(3)
Then smoothening is carried out by convolving input
image with the gradient function.
).(),(),( yxfyxGyxf  (4)
3.2.2. Connected component determination
Connected component analysis is used to find
connected region in the color image as well as in the
Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36
www.ijera.com 35 | P a g e
binary images. There may be character or broken
letters. Each region which is in maximum is called a
connected component.
3.2.3 Closing
The closing operation defined as, image I by a
structuring element T is a dilation followed by
erosion as given in equation (5)
I • T = (I Trot) Trot (5)
The dilation operation uses a structuring element for
probing and expanding the shapes contained in the
input image, and the erosion will removes pixels on
object boundaries. The closing and opening operation
has mainly used for removing morphological noise.
Which means opening remove small objects and
closing removes small holes in a given document.
3.2.4. Y-histogram projection
Y-Histogram is performed on the previous results.
The main aim of y-histogram projection is to find all
possible line segments in the images
IV. Experimental result
The experiments are performed in camera
captured document of textbook, which include 30
images of same viewing angle is considered, which
does not contain broken or missing characters.Figure
2 shows input image. Figure 3 shows the canny edge
detected result. Figure 4 and figure 5 shows
connected components and text area respectively.
Fig.2.input image
Fig.4.connected components
Fig.3.canny edge detection
Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36
www.ijera.com 36 | P a g e
Fig.5.Segmented Text line
V. Conclusion
In this paper, we propose a model for
extracting text line from a warped document captured
by a camera. The proposed method is applied on
grayscale document images and is based on an
adaptive document image binarization and the text
line detection. In preprocessing, adaptive
thresholding method has been used which statistically
examine the intensity values of the local
neighborhood of each pixel. This method is useful
when the text image contain strong illumination-
gradient, then use of the average may prove to be
effective in eliminating the ill influence. Text
detection or extraction from the documents is the
next step before segmentation. Here curled text lines
are detected based on histogram and projection.
Compare to previous method it is less sensitive to
higher level of skew and the orientation.Future work
focuses on flattening of the text line which are
extracted by previous method. Rectification is
method is mainly used for straighten the
document one by one. Also we will concentrate
on reducing the perspective and geometric
distortions.
References
[1] Shijian Lu, Ben M. Chen , C.C. Ko ,A
partition approach for the restoration of
camera images of planar and curled
document,Conference on Image and Vision
Computing , 24, 2006, 837-848.
[2] Syed Saqib Bukhari, Faisal Shafait,
Thomas.M. Breuel,Coupled snakelets for
curled text-line segmentation from warped
document images, 10th Int. Conf. on
Document Analysis and Recognition,
2011,33-53
[3] Partha Pratim Roy ·, Umapada Pal ·, Josep
Lladós,Text line extraction in graphical
documents using background and foreground
information, International Journal on
Document Analysis and Recognition,15,
2011, 227-241.
[4] Vaibhav, Gavali.Multi-Oriented and Text
Line Extraction from documents,
International Journal of Computer Science
and Mobile Computing, 2,2013, 285-293
[5] G. Louloudis, B. Gatos, I. Pratikakis, K.
Halatsis,A Block-Based Hough Transform
Mapping for Text Line Detection in
Handwritten Documents, science direct
Image and Vision Computing . vol. 24, no.
10, Oct. 2010, pp.837-848
[6] Syed Saqib Bukhari and Thomas M.
Breuel,”Text line information extraction from
Grayscale camera-captured document
images”. IEEE Trans. On Pattern Analysis
and Machine Intelligence, vol. 26, Oct. 2011,
pp. 1295-1306
[7] Shafait, F., Keysers, D., Breuel, and
T.M,Performance evaluation and
benchmarking of six page segmentation
algorithms, IEEE Trans.Pattern Anal. Mach.
Intell. Vol.30 pp.941–954 (2008)
[8] Aivind Due Trier and Torfinn
Taxt,Evaluation of Binarization Methods for
Document Images, IEEE transactions on
pattern analysis and machine intelligence,
vol. 3, march 1995
[9] Irina Rabaev, Ofer Biller, Jihad El-Sana,
Klara Kedem,Text Line Detection in
Corrupted and Damaged Historical
Manuscripts, IEEE Transactions on Pattern
Analysis and Machine Intelligence, vol 34,
April 2012, pp.707-722.
[10] E. Kavallieratou, N. Fakotakis, G.
Kokkinakis,“Skew angle estimation for
printed and handwritten documents using the
Wigner–Ville distribution,Image and Vision
Computing” vol .20 , pp.813–824.
[11] U. Pal, B.B. Chaudhuri, “Multi-oriented text
lines detection and their skew estimation”,
Third Indian Conference on Computer
Vision, Graphics and Image Processing,, pp.
270–275

Contenu connexe

Tendances

Text Extraction of Colour Images using Mathematical Morphology & HAAR Transform
Text Extraction of Colour Images using Mathematical Morphology & HAAR TransformText Extraction of Colour Images using Mathematical Morphology & HAAR Transform
Text Extraction of Colour Images using Mathematical Morphology & HAAR TransformIOSR Journals
 
Comparative analysis of c99 and topictiling text segmentation algorithms
Comparative analysis of c99 and topictiling text segmentation algorithmsComparative analysis of c99 and topictiling text segmentation algorithms
Comparative analysis of c99 and topictiling text segmentation algorithmseSAT Journals
 
Comparative analysis of c99 and topictiling text
Comparative analysis of c99 and topictiling textComparative analysis of c99 and topictiling text
Comparative analysis of c99 and topictiling texteSAT Publishing House
 
IRJET- Object Detection using Hausdorff Distance
IRJET-  	  Object Detection using Hausdorff DistanceIRJET-  	  Object Detection using Hausdorff Distance
IRJET- Object Detection using Hausdorff DistanceIRJET Journal
 
Texture features based text extraction from images using DWT and K-means clus...
Texture features based text extraction from images using DWT and K-means clus...Texture features based text extraction from images using DWT and K-means clus...
Texture features based text extraction from images using DWT and K-means clus...Divya Gera
 
An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...
An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...
An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...iosrjce
 
A Survey on Tamil Handwritten Character Recognition using OCR Techniques
A Survey on Tamil Handwritten Character Recognition using OCR TechniquesA Survey on Tamil Handwritten Character Recognition using OCR Techniques
A Survey on Tamil Handwritten Character Recognition using OCR Techniquescscpconf
 
Text Extraction from Image using Python
Text Extraction from Image using PythonText Extraction from Image using Python
Text Extraction from Image using Pythonijtsrd
 
Text extraction from natural scene image, a survey
Text extraction from natural scene image, a surveyText extraction from natural scene image, a survey
Text extraction from natural scene image, a surveySOYEON KIM
 
K2 Algorithm-based Text Detection with An Adaptive Classifier Threshold
K2 Algorithm-based Text Detection with An Adaptive Classifier ThresholdK2 Algorithm-based Text Detection with An Adaptive Classifier Threshold
K2 Algorithm-based Text Detection with An Adaptive Classifier ThresholdCSCJournals
 
Image to text Converter
Image to text ConverterImage to text Converter
Image to text ConverterDhiraj Raj
 
Self-Directing Text Detection and Removal from Images with Smoothing
Self-Directing Text Detection and Removal from Images with SmoothingSelf-Directing Text Detection and Removal from Images with Smoothing
Self-Directing Text Detection and Removal from Images with SmoothingPriyanka Wagh
 
IRJET- Devnagari Text Detection
IRJET- Devnagari Text DetectionIRJET- Devnagari Text Detection
IRJET- Devnagari Text DetectionIRJET Journal
 
SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...
SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...
SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...Cheriyan K M
 
Dimensionality Reduction and Feature Selection Methods for Script Identificat...
Dimensionality Reduction and Feature Selection Methods for Script Identificat...Dimensionality Reduction and Feature Selection Methods for Script Identificat...
Dimensionality Reduction and Feature Selection Methods for Script Identificat...ITIIIndustries
 

Tendances (19)

Text Extraction of Colour Images using Mathematical Morphology & HAAR Transform
Text Extraction of Colour Images using Mathematical Morphology & HAAR TransformText Extraction of Colour Images using Mathematical Morphology & HAAR Transform
Text Extraction of Colour Images using Mathematical Morphology & HAAR Transform
 
Comparative analysis of c99 and topictiling text segmentation algorithms
Comparative analysis of c99 and topictiling text segmentation algorithmsComparative analysis of c99 and topictiling text segmentation algorithms
Comparative analysis of c99 and topictiling text segmentation algorithms
 
Comparative analysis of c99 and topictiling text
Comparative analysis of c99 and topictiling textComparative analysis of c99 and topictiling text
Comparative analysis of c99 and topictiling text
 
IRJET- Object Detection using Hausdorff Distance
IRJET-  	  Object Detection using Hausdorff DistanceIRJET-  	  Object Detection using Hausdorff Distance
IRJET- Object Detection using Hausdorff Distance
 
Texture features based text extraction from images using DWT and K-means clus...
Texture features based text extraction from images using DWT and K-means clus...Texture features based text extraction from images using DWT and K-means clus...
Texture features based text extraction from images using DWT and K-means clus...
 
An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...
An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...
An Efficient Segmentation Technique for Machine Printed Devanagiri Script: Bo...
 
A Survey on Tamil Handwritten Character Recognition using OCR Techniques
A Survey on Tamil Handwritten Character Recognition using OCR TechniquesA Survey on Tamil Handwritten Character Recognition using OCR Techniques
A Survey on Tamil Handwritten Character Recognition using OCR Techniques
 
Text Extraction from Image using Python
Text Extraction from Image using PythonText Extraction from Image using Python
Text Extraction from Image using Python
 
Text extraction from natural scene image, a survey
Text extraction from natural scene image, a surveyText extraction from natural scene image, a survey
Text extraction from natural scene image, a survey
 
K2 Algorithm-based Text Detection with An Adaptive Classifier Threshold
K2 Algorithm-based Text Detection with An Adaptive Classifier ThresholdK2 Algorithm-based Text Detection with An Adaptive Classifier Threshold
K2 Algorithm-based Text Detection with An Adaptive Classifier Threshold
 
Text Detection and Recognition
Text Detection and RecognitionText Detection and Recognition
Text Detection and Recognition
 
Image to text Converter
Image to text ConverterImage to text Converter
Image to text Converter
 
Self-Directing Text Detection and Removal from Images with Smoothing
Self-Directing Text Detection and Removal from Images with SmoothingSelf-Directing Text Detection and Removal from Images with Smoothing
Self-Directing Text Detection and Removal from Images with Smoothing
 
IRJET- Devnagari Text Detection
IRJET- Devnagari Text DetectionIRJET- Devnagari Text Detection
IRJET- Devnagari Text Detection
 
SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...
SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...
SCENE TEXT RECOGNITION IN MOBILE APPLICATION BY CHARACTER DESCRIPTOR AND STRU...
 
C04741319
C04741319C04741319
C04741319
 
Text Mining
Text MiningText Mining
Text Mining
 
40120140501009
4012014050100940120140501009
40120140501009
 
Dimensionality Reduction and Feature Selection Methods for Script Identificat...
Dimensionality Reduction and Feature Selection Methods for Script Identificat...Dimensionality Reduction and Feature Selection Methods for Script Identificat...
Dimensionality Reduction and Feature Selection Methods for Script Identificat...
 

En vedette

La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...
La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...
La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...Universidad Veracruzana
 
20 Ideas for your Website Homepage Content
20 Ideas for your Website Homepage Content20 Ideas for your Website Homepage Content
20 Ideas for your Website Homepage ContentBarry Feldman
 
50 Essential Content Marketing Hacks (Content Marketing World)
50 Essential Content Marketing Hacks (Content Marketing World)50 Essential Content Marketing Hacks (Content Marketing World)
50 Essential Content Marketing Hacks (Content Marketing World)Heinz Marketing Inc
 
Prototyping is an attitude
Prototyping is an attitudePrototyping is an attitude
Prototyping is an attitudeWith Company
 
10 Insightful Quotes On Designing A Better Customer Experience
10 Insightful Quotes On Designing A Better Customer Experience10 Insightful Quotes On Designing A Better Customer Experience
10 Insightful Quotes On Designing A Better Customer ExperienceYuan Wang
 
How to Build a Dynamic Social Media Plan
How to Build a Dynamic Social Media PlanHow to Build a Dynamic Social Media Plan
How to Build a Dynamic Social Media PlanPost Planner
 
SEO: Getting Personal
SEO: Getting PersonalSEO: Getting Personal
SEO: Getting PersonalKirsty Hulse
 

En vedette (8)

La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...
La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...
La documentación hipermedia en el ámbito de la publicidad y las relaciones pú...
 
20 Ideas for your Website Homepage Content
20 Ideas for your Website Homepage Content20 Ideas for your Website Homepage Content
20 Ideas for your Website Homepage Content
 
50 Essential Content Marketing Hacks (Content Marketing World)
50 Essential Content Marketing Hacks (Content Marketing World)50 Essential Content Marketing Hacks (Content Marketing World)
50 Essential Content Marketing Hacks (Content Marketing World)
 
Prototyping is an attitude
Prototyping is an attitudePrototyping is an attitude
Prototyping is an attitude
 
10 Insightful Quotes On Designing A Better Customer Experience
10 Insightful Quotes On Designing A Better Customer Experience10 Insightful Quotes On Designing A Better Customer Experience
10 Insightful Quotes On Designing A Better Customer Experience
 
How to Build a Dynamic Social Media Plan
How to Build a Dynamic Social Media PlanHow to Build a Dynamic Social Media Plan
How to Build a Dynamic Social Media Plan
 
SEO: Getting Personal
SEO: Getting PersonalSEO: Getting Personal
SEO: Getting Personal
 
Succession “Losers”: What Happens to Executives Passed Over for the CEO Job?
Succession “Losers”: What Happens to Executives Passed Over for the CEO Job? Succession “Losers”: What Happens to Executives Passed Over for the CEO Job?
Succession “Losers”: What Happens to Executives Passed Over for the CEO Job?
 

Similaire à F045053236

Enhancement and Segmentation of Historical Records
Enhancement and Segmentation of Historical RecordsEnhancement and Segmentation of Historical Records
Enhancement and Segmentation of Historical Recordscsandit
 
Header Based Classification of Journals Using Document Image Segmentation and...
Header Based Classification of Journals Using Document Image Segmentation and...Header Based Classification of Journals Using Document Image Segmentation and...
Header Based Classification of Journals Using Document Image Segmentation and...CSCJournals
 
Text-Image Separation in Document Images using Boundary/Perimeter Detection
Text-Image Separation in Document Images using Boundary/Perimeter DetectionText-Image Separation in Document Images using Boundary/Perimeter Detection
Text-Image Separation in Document Images using Boundary/Perimeter DetectionIDES Editor
 
Persian arabic document segmentation based on hybrid approach
Persian arabic document segmentation based on hybrid approachPersian arabic document segmentation based on hybrid approach
Persian arabic document segmentation based on hybrid approachijcsa
 
Finding similarities between structured documents as a crucial stage for gene...
Finding similarities between structured documents as a crucial stage for gene...Finding similarities between structured documents as a crucial stage for gene...
Finding similarities between structured documents as a crucial stage for gene...Alexander Decker
 
IRJET - Object Detection using Hausdorff Distance
IRJET -  	  Object Detection using Hausdorff DistanceIRJET -  	  Object Detection using Hausdorff Distance
IRJET - Object Detection using Hausdorff DistanceIRJET Journal
 
Binarization of Degraded Text documents and Palm Leaf Manuscripts
Binarization of Degraded Text documents and Palm Leaf ManuscriptsBinarization of Degraded Text documents and Palm Leaf Manuscripts
Binarization of Degraded Text documents and Palm Leaf ManuscriptsIRJET Journal
 
An effective approach to offline arabic handwriting recognition
An effective approach to offline arabic handwriting recognitionAn effective approach to offline arabic handwriting recognition
An effective approach to offline arabic handwriting recognitionijaia
 
Text Extraction from Image Using GAMMA Correction Method.
Text Extraction from Image Using GAMMA Correction Method.Text Extraction from Image Using GAMMA Correction Method.
Text Extraction from Image Using GAMMA Correction Method.IRJET Journal
 
A Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In VideosA Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In VideosCSCJournals
 
A Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In VideosA Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In VideosCSCJournals
 
Improved wolf algorithm on document images detection using optimum mean techn...
Improved wolf algorithm on document images detection using optimum mean techn...Improved wolf algorithm on document images detection using optimum mean techn...
Improved wolf algorithm on document images detection using optimum mean techn...journalBEEI
 
Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...
Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...
Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...CSCJournals
 
A Run Length Smoothing-Based Algorithm for Non-Manhattan Document Segmentation
A Run Length Smoothing-Based Algorithm for Non-Manhattan Document SegmentationA Run Length Smoothing-Based Algorithm for Non-Manhattan Document Segmentation
A Run Length Smoothing-Based Algorithm for Non-Manhattan Document SegmentationUniversity of Bari (Italy)
 
Proposed technique-for-edge-matching-of-torn-paper
Proposed technique-for-edge-matching-of-torn-paperProposed technique-for-edge-matching-of-torn-paper
Proposed technique-for-edge-matching-of-torn-paperEditor IJMTER
 
Cc31331335
Cc31331335Cc31331335
Cc31331335IJMER
 
Optical Character Recognition from Text Image
Optical Character Recognition from Text ImageOptical Character Recognition from Text Image
Optical Character Recognition from Text ImageEditor IJCATR
 
Texture based feature extraction and object tracking
Texture based feature extraction and object trackingTexture based feature extraction and object tracking
Texture based feature extraction and object trackingPriyanka Goswami
 
A Texture Based Methodology for Text Region Extraction from Low Resolution Na...
A Texture Based Methodology for Text Region Extraction from Low Resolution Na...A Texture Based Methodology for Text Region Extraction from Low Resolution Na...
A Texture Based Methodology for Text Region Extraction from Low Resolution Na...CSCJournals
 

Similaire à F045053236 (20)

Enhancement and Segmentation of Historical Records
Enhancement and Segmentation of Historical RecordsEnhancement and Segmentation of Historical Records
Enhancement and Segmentation of Historical Records
 
Header Based Classification of Journals Using Document Image Segmentation and...
Header Based Classification of Journals Using Document Image Segmentation and...Header Based Classification of Journals Using Document Image Segmentation and...
Header Based Classification of Journals Using Document Image Segmentation and...
 
Text-Image Separation in Document Images using Boundary/Perimeter Detection
Text-Image Separation in Document Images using Boundary/Perimeter DetectionText-Image Separation in Document Images using Boundary/Perimeter Detection
Text-Image Separation in Document Images using Boundary/Perimeter Detection
 
Persian arabic document segmentation based on hybrid approach
Persian arabic document segmentation based on hybrid approachPersian arabic document segmentation based on hybrid approach
Persian arabic document segmentation based on hybrid approach
 
Finding similarities between structured documents as a crucial stage for gene...
Finding similarities between structured documents as a crucial stage for gene...Finding similarities between structured documents as a crucial stage for gene...
Finding similarities between structured documents as a crucial stage for gene...
 
IRJET - Object Detection using Hausdorff Distance
IRJET -  	  Object Detection using Hausdorff DistanceIRJET -  	  Object Detection using Hausdorff Distance
IRJET - Object Detection using Hausdorff Distance
 
Binarization of Degraded Text documents and Palm Leaf Manuscripts
Binarization of Degraded Text documents and Palm Leaf ManuscriptsBinarization of Degraded Text documents and Palm Leaf Manuscripts
Binarization of Degraded Text documents and Palm Leaf Manuscripts
 
An effective approach to offline arabic handwriting recognition
An effective approach to offline arabic handwriting recognitionAn effective approach to offline arabic handwriting recognition
An effective approach to offline arabic handwriting recognition
 
Text Extraction from Image Using GAMMA Correction Method.
Text Extraction from Image Using GAMMA Correction Method.Text Extraction from Image Using GAMMA Correction Method.
Text Extraction from Image Using GAMMA Correction Method.
 
A Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In VideosA Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In Videos
 
A Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In VideosA Survey On Thresholding Operators of Text Extraction In Videos
A Survey On Thresholding Operators of Text Extraction In Videos
 
E1803012329
E1803012329E1803012329
E1803012329
 
Improved wolf algorithm on document images detection using optimum mean techn...
Improved wolf algorithm on document images detection using optimum mean techn...Improved wolf algorithm on document images detection using optimum mean techn...
Improved wolf algorithm on document images detection using optimum mean techn...
 
Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...
Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...
Separation of mixed Document Images in Farsi Scanned Documents Using Blind So...
 
A Run Length Smoothing-Based Algorithm for Non-Manhattan Document Segmentation
A Run Length Smoothing-Based Algorithm for Non-Manhattan Document SegmentationA Run Length Smoothing-Based Algorithm for Non-Manhattan Document Segmentation
A Run Length Smoothing-Based Algorithm for Non-Manhattan Document Segmentation
 
Proposed technique-for-edge-matching-of-torn-paper
Proposed technique-for-edge-matching-of-torn-paperProposed technique-for-edge-matching-of-torn-paper
Proposed technique-for-edge-matching-of-torn-paper
 
Cc31331335
Cc31331335Cc31331335
Cc31331335
 
Optical Character Recognition from Text Image
Optical Character Recognition from Text ImageOptical Character Recognition from Text Image
Optical Character Recognition from Text Image
 
Texture based feature extraction and object tracking
Texture based feature extraction and object trackingTexture based feature extraction and object tracking
Texture based feature extraction and object tracking
 
A Texture Based Methodology for Text Region Extraction from Low Resolution Na...
A Texture Based Methodology for Text Region Extraction from Low Resolution Na...A Texture Based Methodology for Text Region Extraction from Low Resolution Na...
A Texture Based Methodology for Text Region Extraction from Low Resolution Na...
 

Dernier

KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostMatt Ray
 
UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1DianaGray10
 
9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding TeamAdam Moalla
 
Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Adtran
 
NIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 WorkshopNIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 WorkshopBachir Benyammi
 
Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024D Cloud Solutions
 
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UbiTrack UK
 
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...DianaGray10
 
Building AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxBuilding AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxUdaiappa Ramachandran
 
Machine Learning Model Validation (Aijun Zhang 2024).pdf
Machine Learning Model Validation (Aijun Zhang 2024).pdfMachine Learning Model Validation (Aijun Zhang 2024).pdf
Machine Learning Model Validation (Aijun Zhang 2024).pdfAijun Zhang
 
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...Aggregage
 
Empowering Africa's Next Generation: The AI Leadership Blueprint
Empowering Africa's Next Generation: The AI Leadership BlueprintEmpowering Africa's Next Generation: The AI Leadership Blueprint
Empowering Africa's Next Generation: The AI Leadership BlueprintMahmoud Rabie
 
Comparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioComparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioChristian Posta
 
20230202 - Introduction to tis-py
20230202 - Introduction to tis-py20230202 - Introduction to tis-py
20230202 - Introduction to tis-pyJamie (Taka) Wang
 
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IES VE
 
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAAnypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAshyamraj55
 
COMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a WebsiteCOMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a Websitedgelyza
 
OpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureOpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureEric D. Schabell
 

Dernier (20)

KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCostKubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
KubeConEU24-Monitoring Kubernetes and Cloud Spend with OpenCost
 
UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1UiPath Platform: The Backend Engine Powering Your Automation - Session 1
UiPath Platform: The Backend Engine Powering Your Automation - Session 1
 
9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team9 Steps For Building Winning Founding Team
9 Steps For Building Winning Founding Team
 
Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™Meet the new FSP 3000 M-Flex800™
Meet the new FSP 3000 M-Flex800™
 
NIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 WorkshopNIST Cybersecurity Framework (CSF) 2.0 Workshop
NIST Cybersecurity Framework (CSF) 2.0 Workshop
 
Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024Artificial Intelligence & SEO Trends for 2024
Artificial Intelligence & SEO Trends for 2024
 
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
UWB Technology for Enhanced Indoor and Outdoor Positioning in Physiological M...
 
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
Connector Corner: Extending LLM automation use cases with UiPath GenAI connec...
 
Building AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptxBuilding AI-Driven Apps Using Semantic Kernel.pptx
Building AI-Driven Apps Using Semantic Kernel.pptx
 
20230104 - machine vision
20230104 - machine vision20230104 - machine vision
20230104 - machine vision
 
201610817 - edge part1
201610817 - edge part1201610817 - edge part1
201610817 - edge part1
 
Machine Learning Model Validation (Aijun Zhang 2024).pdf
Machine Learning Model Validation (Aijun Zhang 2024).pdfMachine Learning Model Validation (Aijun Zhang 2024).pdf
Machine Learning Model Validation (Aijun Zhang 2024).pdf
 
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
The Data Metaverse: Unpacking the Roles, Use Cases, and Tech Trends in Data a...
 
Empowering Africa's Next Generation: The AI Leadership Blueprint
Empowering Africa's Next Generation: The AI Leadership BlueprintEmpowering Africa's Next Generation: The AI Leadership Blueprint
Empowering Africa's Next Generation: The AI Leadership Blueprint
 
Comparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and IstioComparing Sidecar-less Service Mesh from Cilium and Istio
Comparing Sidecar-less Service Mesh from Cilium and Istio
 
20230202 - Introduction to tis-py
20230202 - Introduction to tis-py20230202 - Introduction to tis-py
20230202 - Introduction to tis-py
 
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
IESVE Software for Florida Code Compliance Using ASHRAE 90.1-2019
 
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPAAnypoint Code Builder , Google Pub sub connector and MuleSoft RPA
Anypoint Code Builder , Google Pub sub connector and MuleSoft RPA
 
COMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a WebsiteCOMPUTER 10 Lesson 8 - Building a Website
COMPUTER 10 Lesson 8 - Building a Website
 
OpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability AdventureOpenShift Commons Paris - Choose Your Own Observability Adventure
OpenShift Commons Paris - Choose Your Own Observability Adventure
 

F045053236

  • 1. Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36 www.ijera.com 32 | P a g e S Text line Segmentation of Curved Document Images Anusree.M* , Dhanya.M.Dhanalakshmy** * (Department of Computer Science, Amrita Vishwa Vidhyapeetham, Coimbatore -641 112) ** (Department of Computer Science, Amrita Vishwa Vidhyapeetham, Coimbatore -641 112) ABSTRACT Document image analysis has been widely used in historical and heritage studies, education and digital library. Document image analytical techniques are mainly used for improving the human readability and the OCR quality of the document. During the digitization, camera captured images contain warped document due perspective and geometric distortions. The main difficulty is text line detection in the document. Many algorithms had been proposed to address the problem of printed document text line detection, but they failed to extract text lines in curved document. This paper describes a segmentation technique that detects the curled text line in camera captured document images. Keywords – Curved document images, Gradient vector flow, Gray scale image, Optical character recognition, Rectification. I. Introduction Digitization of document is the most commonly used method for increasing the quality and compatibility of the document. Now a days, compared to scanned document, camera captured document has been widely used for digitization. Most of the camera captured documents such as degraded books and newspapers suffer curling of text lines due to perspective distortion. There is a huge amount of document images in libraries and in various national archives that have not been exploited electronically. If these documents need to be electronically available, the first and major step will be document segmentation into text lines. Before text line segmentation, preprocessing steps are carried out. Thresholding is used for the binarization that converts a gray-scale (8 bit per pixel) document image into a binary document image (1 bit per pixel). It can be divided into global thresholding and local thresholding. For flattening or straightening the document, first stage is to segment the text line from the document images. This is the most difficult step in rectification method. There are many methods in text line extraction in the typical text line extraction stage. But in the case of the curled document image, it is difficult to use usual algorithms such as projection profile, x cut, run length smearing etc. Here we are proposing method text line segmentation for curved document images. II. Literature survey There are many methods to extract text lines from single-oriented documents. Shijian Lu et al. [1] proposed a method for implementing the text line segmentation of curved document. The rectification of curled text lines has been carried out by using partition based approaches. Partition based approaches include baseline detection, x line, and vertical stroke identification. Here x line and baseline have been estimated using least square fitting of the character tip point. Documents are partitioned into multiple quadrilateral patches. Finally target rectangle has been constructed. It is done using aspect ratio and number of the characters enclosed within the partitioned block. This method work on Latin-based document and the authors classified the document in to six categories based on shape such as character span, character ascender and descender. Compared with the existing algorithms, this method doesn’t need any special hardware or camera calibration. But the main disadvantage of this system is that it will not work on low resolution images. This method is not suitable for handwritten and historical documents. Syed Saqib Bukhari et al. [2] presented a new algorithm for text line segmentation of curled document images. Here curled text-line segmentation is performed by adapting active contour (snake) which has been used for de-wrapping and improving the OCR quality as well as the readability. Snakes are used for estimating the local pair of x line and base line. This techniques uses x line and base line estimation using connected component analysis and this is applied to the monocular single camera captured document images. The overlapping x line and baseline are considered as segmented text lines. Here coupled snakes model has been used instead of snake model. The snake moves towards targeted object under the influence of internal energy and external energy. The internal and external energy has been calculated from image content and snake point RESEARCH ARTICLE OPEN ACCESS
  • 2. Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36 www.ijera.com 33 | P a g e respectively. After finding the external energy, the evolving and deformation of the snakes have to be carried out using gradient vector flow. The main advantage of this method is that it is less sensitive to high degree of curl and skew, while the challenge is that it produces many number of over-and under- segmentation errors. Partha Pratim Roy et al. [3] proposed text line extraction using foreground and background information in the document image. For the extraction of curved document images, water reservoir technique has been used. Using positional information and the size, each individual component is grouped together as a cluster. In the second stage, the potential region has been found and individual lines are estimated using candidate region. The main advantage of this method is to determine the document containing multi-oriented and curled lines. The steps include initial character clustering, grouping cluster, candidate point extraction followed by extension of cluster group. The efficiency of the algorithm depends upon the binarization result. So it does not work with the colored document. But it is independent of font style, rotation, scale etc.. Vaibhav Gavali et al. [4] describe the algorithm that is independent of the text and the color of the document. The paper describes a technique which extracts text lines from a curved gray scale images. The gray scale image has been taken as input; binarized and discrete haar wavelet is applied followed by thresholding to remove some non-text region in the document. The morphological operation has been used for each band to connect disconnected characters. At last, 3 dilated bands are combined using logical AND operator to get segmented text lines. Here 2D haar wavelets are used for edge detection which decomposes the input document in to LL, LH, HL, and HH. The detailed information of the text lines are presents in the LH, HL, HH and this output is threshold using dynamic threshold. G. Louloudis et al. [5] used a block-Based Hough Transform for text line extraction. This method is used for unconstrained handwritten documents. In the first step, preprocessing for enhancement of image, connected component extraction and average character height estimation were carried out. In binarization, adaptive thresholding method has been used. From the result all connected component has been identified and area of the bounding box is calculated. The proposed approach consists of partitioning the connected component space into three subsets, and the splitting of connected components into equally spaced blocks each of them voting in the Hough domain. This proposed method is applied to handwritten document which are in the Latin format. The output of the document is post- processed to reduce more than one line which corresponds to same Hough domain. III. Proposed method The architecture of the system is shown in the figure 1. In the first stage, input color image is converted to gray scale and preprocessed. The next stage determines edges. At last, curved text lines are segmented using morphological operation. Input image Segmented text line Fig.1.Proposed system architecture 3.1. Pre-processing Pre-processing is the first step for document image analysis. Here pre-processing include binarization and noise reduction. 3.1.1. Binarization In an OCR, one of the main processing stages is binarization of document images. Binarization of a text image should give the foreground text in black and noisy background in white. There are different thresholding methods already exist; they don’t give exact results for all types of document images. Few algorithms might work better for one type of documents while they might give poor results for other types where there is a complex background, extremely low intensity variation. Depending upon the document the output of threshold will vary, hence in this project different methods were tried for getting better result. 3.1.1.1. Otsu’s method Otsu’s method is one of the famous global thresholding methods. This method tries to find the threshold T which separates the gray-level histogram in an optimal way into two segments. Here the calculation of inter-classes or intra-classes variances is based on the normalized histogram of the image H= [h0...h255] where Σ hi=1. In Otsu algorithm, result has noise in the form of background being detected as foreground. 3.1.1.2. Niblack’s algorithm Niblack’s algorithm calculates a pixel-wise threshold by sliding a rectangular window over the gray level image. The computation of threshold is based on the local mean m and the standard deviation s of all the pixels in the window and is given by the equation 1 below Preprocessing Text line segmentation
  • 3. Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36 www.ijera.com 34 | P a g e skmT  )1( Here m is the average value of the pixels pi, and k is fixed to -0.2. Advantage of Niblack is that it always identifies the text regions correctly as foreground but on the other hand tends to produce a large amount of binarization noise in non-text regions also. 3.1.1.3. Savola algorithm Savola algorithm claims to improve Niblack’s method by computing the threshold using the dynamic range of image gray-value standard deviation.The binarization is given by equation (2) ) R σ k)(1m(1T  )2( Where k is se to 0.5 and R varies from 0 to 255. This method outperforms Niblack’s algorithm in images where the text pixels have near 0 gray-values and the background pixels have near 255 gray-values. However, in images the gray values of text and non- text pixels are close to each other results degrade significantly. Thresholding is used to segment an image by setting all pixels whose intensity values are above a threshold to a foreground value and all the remaining pixels to a background value. The traditional thresholding operator uses a global threshold for entire images; this method changes the threshold dynamically over the image. it can accommodate changing lighting conditions in the image, such as those occurring as a result of a strong illumination gradient or shadows. 3.1.1.4. Adaptive thresholding Adaptive thresholding takes a grayscale or color image as input and it produces a binary image representing the text lines in the image. The pixel value is above threshold, it is set to the foreground value; otherwise it assumes the background value. Local adaptive thresholding will select an individual threshold for each pixel based on the range of intensity values in its local neighborhood pixels. This allows for binarization of a document whose global intensity histogram doesn't contain distinctive peaks. The algorithm involves following steps: 1) Convolve the image with a suitable statistical operator. 2) Subtract the original from the convolved image. 3) Threshold the difference image with C. 4) Invert the threshold image. Out of the four methods described above, adaptive binarization algorithm was found to give better results for our dataset. 3.1.2 Noise Removal Morphological image processing is a set of operations related to the shape or morphology of features in an image. Here median filter is mainly used for marginal noise reduction. The median filter is mainly used for removal of salt and pepper noise. It replaces value of the pixel by the median of the intensity values in the neighborhood of that pixel. 3.2. Text line detection Optical character recognition (OCR) or text recognition is an important area in commercial software development and academic research. The application areas of digital document processing are office and library automation, publishing houses, communication technology, banking sector, historical and many other domain areas. A large number of techniques have been proposed to address the problem of document text extraction. Text line detection is a critical stage towards unconstrained handwritten document recognition. It refers to the segmentation of a document page image into distinct entities, the text lines. The challenges in problems that appear in this stage are the difference in the skew angle between lines on the page, overlapping words and adjacent lines touching. 3.2.1. Canny Edge Detection To detect the edges in the document canny edge detector has been used .The main advantages this method are it has low error rate, edge point is localized and accurate result is obtained on the single edge point pixel. The canny detector algorithm involves four steps:- 1. Smoothing the input image using guasssain filter. 2. Find out gradient magnitude and angle 3. Nonmaxima suppreeion to the previous result. 4. Apply double thresholding to identify the edges. The image smoothening is carried out by using circular 2D Gaussian function which is denoted by equation 3:- 2 22 2 )( ).(  yx eyxG   (3) Then smoothening is carried out by convolving input image with the gradient function. ).(),(),( yxfyxGyxf  (4) 3.2.2. Connected component determination Connected component analysis is used to find connected region in the color image as well as in the
  • 4. Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36 www.ijera.com 35 | P a g e binary images. There may be character or broken letters. Each region which is in maximum is called a connected component. 3.2.3 Closing The closing operation defined as, image I by a structuring element T is a dilation followed by erosion as given in equation (5) I • T = (I Trot) Trot (5) The dilation operation uses a structuring element for probing and expanding the shapes contained in the input image, and the erosion will removes pixels on object boundaries. The closing and opening operation has mainly used for removing morphological noise. Which means opening remove small objects and closing removes small holes in a given document. 3.2.4. Y-histogram projection Y-Histogram is performed on the previous results. The main aim of y-histogram projection is to find all possible line segments in the images IV. Experimental result The experiments are performed in camera captured document of textbook, which include 30 images of same viewing angle is considered, which does not contain broken or missing characters.Figure 2 shows input image. Figure 3 shows the canny edge detected result. Figure 4 and figure 5 shows connected components and text area respectively. Fig.2.input image Fig.4.connected components Fig.3.canny edge detection
  • 5. Anusree.M et al Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 4, Issue 5( Version 5), May 2014, pp.32-36 www.ijera.com 36 | P a g e Fig.5.Segmented Text line V. Conclusion In this paper, we propose a model for extracting text line from a warped document captured by a camera. The proposed method is applied on grayscale document images and is based on an adaptive document image binarization and the text line detection. In preprocessing, adaptive thresholding method has been used which statistically examine the intensity values of the local neighborhood of each pixel. This method is useful when the text image contain strong illumination- gradient, then use of the average may prove to be effective in eliminating the ill influence. Text detection or extraction from the documents is the next step before segmentation. Here curled text lines are detected based on histogram and projection. Compare to previous method it is less sensitive to higher level of skew and the orientation.Future work focuses on flattening of the text line which are extracted by previous method. Rectification is method is mainly used for straighten the document one by one. Also we will concentrate on reducing the perspective and geometric distortions. References [1] Shijian Lu, Ben M. Chen , C.C. Ko ,A partition approach for the restoration of camera images of planar and curled document,Conference on Image and Vision Computing , 24, 2006, 837-848. [2] Syed Saqib Bukhari, Faisal Shafait, Thomas.M. Breuel,Coupled snakelets for curled text-line segmentation from warped document images, 10th Int. Conf. on Document Analysis and Recognition, 2011,33-53 [3] Partha Pratim Roy ·, Umapada Pal ·, Josep Lladós,Text line extraction in graphical documents using background and foreground information, International Journal on Document Analysis and Recognition,15, 2011, 227-241. [4] Vaibhav, Gavali.Multi-Oriented and Text Line Extraction from documents, International Journal of Computer Science and Mobile Computing, 2,2013, 285-293 [5] G. Louloudis, B. Gatos, I. Pratikakis, K. Halatsis,A Block-Based Hough Transform Mapping for Text Line Detection in Handwritten Documents, science direct Image and Vision Computing . vol. 24, no. 10, Oct. 2010, pp.837-848 [6] Syed Saqib Bukhari and Thomas M. Breuel,”Text line information extraction from Grayscale camera-captured document images”. IEEE Trans. On Pattern Analysis and Machine Intelligence, vol. 26, Oct. 2011, pp. 1295-1306 [7] Shafait, F., Keysers, D., Breuel, and T.M,Performance evaluation and benchmarking of six page segmentation algorithms, IEEE Trans.Pattern Anal. Mach. Intell. Vol.30 pp.941–954 (2008) [8] Aivind Due Trier and Torfinn Taxt,Evaluation of Binarization Methods for Document Images, IEEE transactions on pattern analysis and machine intelligence, vol. 3, march 1995 [9] Irina Rabaev, Ofer Biller, Jihad El-Sana, Klara Kedem,Text Line Detection in Corrupted and Damaged Historical Manuscripts, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol 34, April 2012, pp.707-722. [10] E. Kavallieratou, N. Fakotakis, G. Kokkinakis,“Skew angle estimation for printed and handwritten documents using the Wigner–Ville distribution,Image and Vision Computing” vol .20 , pp.813–824. [11] U. Pal, B.B. Chaudhuri, “Multi-oriented text lines detection and their skew estimation”, Third Indian Conference on Computer Vision, Graphics and Image Processing,, pp. 270–275