SlideShare a Scribd company logo
1 of 55
Guided By : Prof. Vineet Garg.

Presented By: Sulabh Pal,Suresh Kumar, Himanshu Srivastava.

Date: April 27th 2012.
• Section I (Introduction)
•   Historical Background
                                          Sulabh
•   Ant System
                                          Pal



• Section II
•   The proposed scheme                   Himanshu
•   Experimental results                  Srivastava


• Section III (Application+Conclusions)
•   Application
•   Conclusions and
•   References                            Suresh
                                          Kumar
Introduction
• Discrete optimization problems difficult to solve

• “Soft computing techniques” developed in past ten
  years:
  o   Genetic algorithms (GAs)
        based on natural selection and genetics
  o   Ant Colony Optimization (ACO)
        modeling ant colony behavior
• Optimization Technique Proposed by Marco
 Dorigo in the early ’90

• Often applied to TSP (Travelling Salesman
  Problem): shortest path between n nodes
• A stochastic construction procedure
• Probabilistically build a solution
• Iteratively adding solution components to partial
   solutions
- Heuristic information
- Pheromone trail
 • Reinforcement Learning reminiscence
 • Modify the problem representation at each
   iteration
• Ants work concurrently and independently
• Collective interaction via indirect communication
  leads to good solutions
Basic Phenomenon of Ant For
              Finding Food
Random Way To Find The Way
• The pheromone
                                concentration on trail B will
                                increase at a higher rate
                                than on A, and soon the
                                ants on route A will choose
• Since the route B is          to follow route B
  shorter, the ants on this   • Since most ants will no
  path will complete the        longer travel on route A, and
  travel more times and         since the pheromone is
  thereby lay more              volatile, trail A will start
  pheromone over it.            evaporating
                              • Only the shortest route will
                                remain!
• (a) The initial graph with distances.
• (b) At time t = 0 there is no trail on the graph edges.
• (c) At time t = 1 trail is stronger on shorter edges.
• Ants (blind) navigate from nest to food source
• Shortest path is discovered via pheromone trails
  o each ant moves at random
  o pheromone is deposited on path
  o ants detect lead ant’s path, inclined to follow
  o more pheromone on path increases probability of
    path being followed
• Starting node selected at random
• Path selected at random
  o based on amount of “trail” present on possible paths
    from starting node
  o higher probability for paths with more “trail”
• Ant reaches next node, selects next path
• Continues until reaches starting node
• Finished “tour” is a solution
• A completed tour is analyzed for optimality
• “Trail” amount adjusted to favor better solutions
  o better solutions receive more trail
  o worse solutions receive less trail
  o higher probability of ant selecting path that is part of a
    better-performing tour
• New cycle is performed
• Repeated until most ants select the same tour
  on every cycle (convergence to solution)
• Algorithm in Pseudocode:
  o   Initialize Trail
  o   Do While (Stopping Criteria Not Satisfied) – Cycle Loop
          Do Until (Each Ant Completes a Tour) – Tour Loop
          Local Trail Update
          End Do
          Analyze Tours
          Global Trail Update
  o   End Do
NATURE
                     COMPUTER SCIENCE
NATURAL HABITAT      GRAPH(NODES AND EDGES)


NEST AND FOOD        NODES IN GRAPH;START AND
                     DESTINATION
ANTS                 AGENTS; OUR ARTIFICIAL
                     AGENTS
VISIBILITY           THE RECIPROCAL OF
                     DISTANCE
PHEROMONES           ARTIFICIAL PHEROMONES


FORAGING BEHAVIOUR   Random walk through graph
                     (guided by pheromones)
    Scheme:
    Construct ant solutions
    Define attractiveness τ, based on experience from previous
    solutions
    Define specific visibility function, η, for a given problem (e.g.
    distance)
   Ant walk
    Initialize ants and nodes (states)
    Choose next edge probabilistically according to the
    attractiveness and visibility
    Each ant maintains a tabu list of infeasible transitions for
    that iteration
    Update attractiveness of an edge according to the number of
    ants that pass
   through
 Pheromone update
 Parameter is called evaporation rate
 Pheromones = long-term memory of an ant colony
 ρ small low evaporation slow adaptation
 ρ large high evaporation fast adaptation
 Note: rules are probabilistic, so mistakes can be
  made!
 “new pheromone” or Δτ usually contains the base
 attractiveness constant Q and a factor that you want to
 optimize
 (e.g. ) Q/length of tour
Path Selection Process
The proposed scheme
Experimental results
Edge                         The
                 detection                    proposed
                                              ACO



Original image               Edge image         The compensatory edge image




                                          The improved result
(Initialization)




  (a). Original image   (b). Traditional edge   (c). Placing ants on all endpoints
                        detection of (a).
0≦q≦1                  If q≦q0
                                                q0=0.9
                                                q=0.5 ≦q0
                                                q=0.3 ≦q0              otherwise
                                                               1   2
(Walk)
  Pixel=1     Pixel=1     Pixel=1     Pixel=1      Pixel=1




  Pixel=100   Pixel=100   Pixel=1     Pixel=100    Pixel=100




  Pixel=100   Pixel=100   Pixel=1     Pixel=100    Pixel=100



G=24,V=0.2 G=74,V=74
  Pixel=100   Pixel=1     Pixel=100   Pixel=100    Pixel=100




  Pixel=1     Pixel=100   Pixel=100   Pixel=100    Pixel=100
0≦q≦1                         If q≦q0
                                               q0=0.9
                                               q=0.95≧q0
                                               q=0.55≦q0                     otherwise


(Walk)
 Pixel=1     Pixel=1     Pixel=1     Pixel=1       Pixel=1



                                                               0.02     0.000001
 Pixel=100   Pixel=100   Pixel=1     Pixel=100     Pixel=100

                                                                      0.03

 Pixel=100   Pixel=100   Pixel=1     Pixel=100     Pixel=100

                                               High prob

 Pixel=100   Pixel=1     Pixel=100   Pixel=100     Pixel=100




 Pixel=1     Pixel=100   Pixel=100   Pixel=100     Pixel=100
(Stop)




  (a). Two ants meet face to face (b). The ant touches the other’s track. (c). Blind alley




          (d). Measurement of track density.
a. Original image   c. 200 iterations   e. =0.9, =0.8      g. 200 iterations
                                                             =0.9, =0.8




b. Sobel edge detector d. Combine b, c     f. Combine b, e    d. Combine b, g
a. Original image   b. Canny edge detector   d. The improved result from
                                             (b) with 160 iterations.
a. Original image    b. Canny edge detector




c. Traditional ACO   d. The proposed approach
ROUTING IN COMM. NETWORKS

Routing task is performed by Routers.
Routers use “Routing Tables” to direct the data.
                                        If your destination is node 5
                                                next node to 3
                    4
        6

                            3       1




                5               2
                        2
ROUTING IN COMM. NETWORKS

Problem statement
 • Dynamic Routing
At any moment the pathway of a message must be as
small as possible. (Traffic conditions and the structure of
the network are constantly changing)
 • Load balancing
Distribute the changing load over the system and
minimize lost calls.
Computation time comparison per 20 iterations.
Comparision Between ACO and Genetic Algorithm
Application
Application
•   TSP
•   QAP
•   Scheduling
•   VRP
•   Telecommunication Network
•   Graph Coloring
•   Water Distribution Network
•   Swarm Intelligence
•   etc
Example Regarding Swarm Algorithm
URBAN SOLID WASTE
COLLECTION          TRAFFIC FLOW OPTIMIZATION
• Positive Feedback accounts for rapid discovery
  of good solutions
• Distributed computation avoids premature
  convergence
• The greedy heuristic helps find acceptable
  solution in the early solution in the early stages
  of the search process.
• The collective interaction of a population of
  agents.
• Slower convergence than other Heuristics
• Performed poorly for TSP problems larger than 75
  cities.
• No centralized processor to guide the AS towards
  good solutions
  Can solve certain NP-Hard problems in Polynomial
  time
 Directed-Random Search
 Allows a balance between using previous knowledge
  and exploring new
 solutions
 Positive feedback for good solutions/Negative
  feedback for
 bad solutions
 Approximately convergent
 Optimal if not absolutely correct solutions
 In certain examples of ACO, no one “ant”
 is required to actually complete an accurate
 solution
  Problem specific
 Limited to problems that can be simulated by
  graphs and optimized
 Coding difficulties for different problems
 Ineffective utilization of previously acquired
  information,
 specifically the global solution
 Depending on the design of the algorithm, it
  can
 converge towards a (less optimal) solution.
  We might like to add factors to minimize the
  time it takes
 to reach an acceptable solution.
 Use the elements of previous solutions
 This allows for faster convergence
 As we construct more and more
  solutions, there is more
 information available about the probable
  “right” choices to make
 The decision making process might weigh
  exploration vs.
 heuristic value
 Ant System: what we just went over
 Ant Colony System:
 Pseudo-random proportion rule:
 at each decision point for an ant, it has a
  probability (1-q0) of using the
 same probability function as in the Ant
  System or q0 of picking the best
 next node based on previous solutions
 Global Trail Update: only the
 best solution since the start of
 the computation will globally
 update its pheromones
 Local Trail Update: all ants
 consume/decrease
 pheromones along the path
 that they travel
 Elitist Ant System:
 Both the global solution and each   ant update
  their edges with
 pheromones on each iteration
 Applications
• The proposed method efficiently improves edge
  images detected by traditional approaches like
  Sobel or Canny.
• The computation time of the proposed method is
  shorter than traditional ACO.
•   Dorigo M. and G. Di Caro (1999). The Ant Colony Optimization Meta-Heuristic. In D. Corne, M. Dorigo
    and F. Glover, editors, New Ideas in Optimization, McGraw-Hill, 11-32.
•   M. Dorigo and L. M. Gambardella. Ant colonies for the traveling salesman problem. BioSystems, 43:73–
    81, 1997.
•   M. Dorigo and L. M. Gambardella. Ant Colony System: A cooperative learning approach to the traveling
    salesman problem. IEEE Transactions on Evolutionary Computation, 1(1):53–66, 1997.
•   G. Di Caro and M. Dorigo. Mobile agents for adaptive routing. In H. El-Rewini, editor, Proceedings of the 31st
    International Conference on System Sciences (HICSS-31), pages 74–83. IEEE Computer Society Press, Los
    Alamitos, CA, 1998.
•   M. Dorigo, V. Maniezzo, and A. Colorni. The Ant System: An autocatalytic optimizing process. Technical
    Report 91-016 Revised, Dipartimento di Elettronica,Politecnico di Milano, Italy, 1991.
•   L. M. Gambardella, ` E. D. Taillard, and G. Agazzi. MACS-VRPTW: A multiple ant colony system for vehicle
    routing problems with time windows. In D. Corne, M. Dorigo, and F. Glover, editors, New Ideas in
    Optimization, pages 63–76. McGraw Hill, London, UK, 1999.
•   L. M. Gambardella, ` E. D. Taillard, and M. Dorigo. Ant colonies for the quadratic assignment problem.
    Journal of the Operational Research Society,50(2):167–176, 1999.
•   V. Maniezzo and A. Colorni. The Ant System applied to the quadratic assignment problem. IEEE
    Transactions on Data and Knowledge Engineering, 11(5):769–778, 1999.
•   Gambardella L. M., E. Taillard and M. Dorigo (1999). Ant Colonies for the Quadratic Assignment
    Problem. Journal of the Operational Research Society, 50:167-176.
•   D. Martens, M. D. Backer, R. Haesen, J. Vanthienen, M. Snoeck, and B.
    Baesens, “Classification with ant colony optimization,” IEEE Trans. on Evolutionary
    Computation, vol. 11, pp. 651–665, Oct. 2007.
•   R. S. Parpinelli, H. S. Lopes, and A. A. Freitas, “Data mining with an ant colony optimization
    algorithm,” IEEE Trans. on Evolutionary Computation, vol. 6, pp. 321–332, Aug. 2002.
•   S. Ouadfel and M. Batouche, “Ant colony system with local search for Markov random field
    image segmentation,” in Proc. IEEE Int. Conf.on Image Processing, Barcelona, Spain, Sep.
    2003, pp. 133–136.
•   S. L. Hegarat-Mascle, A. Kallel, and X. Descombes, “Ant colony optimization for image
    regularization based on a nonstationary Markov modeling,” IEEE Trans. on Image
    Processing, vol. 16, pp. 865–878, Mar. 2007.
•   A. T. Ghanbarian, E. Kabir, and N. M. Charkari, “Color reduction based on ant colony,” Pattern
    Recognition Letters, vol. 28, pp. 1383–1390, Sep. 2007.
•   A. R. Malisia and H. R. Tizhoosh, “Image thresholding using ant colony optimization,” in Proc.
    Canadian Conf. on Computer and Robot Vision, Quebec, Canada, Jun. 2006, pp. 26–26.
•   R. C. Gonzalez and R. E. Woods, Digital image processing. Harlow: Prentice Hall, 2007.
•   H. Nezamabadi-Pour, S. Saryazdi, and E. Rashedi, “Edge detection using ant algorithms,” Soft
    Computing, vol. 10, pp. 623–628, May 2006.
•   D.-S. Lu and C.-C. Chen, “Edge detection improvement by ant colony optimization,” Pattern
    Recognition Letters, vol. 29, pp. 416–425, Mar.2008.
Thank you for your attention!

More Related Content

What's hot

Solving Quadratic Assignment Problems (QAP) using Ant Colony System
Solving Quadratic Assignment Problems (QAP) using Ant Colony SystemSolving Quadratic Assignment Problems (QAP) using Ant Colony System
Solving Quadratic Assignment Problems (QAP) using Ant Colony SystemAjay Bidyarthy
 
Lecture 9 aco
Lecture 9 acoLecture 9 aco
Lecture 9 acomcradc
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimizationJoy Dutta
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimizationUnnitaDas
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimizationvk1dadhich
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimizationMeenakshi Devi
 
Ant Colony Optimization
Ant Colony OptimizationAnt Colony Optimization
Ant Colony OptimizationPratik Poddar
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimizationAbdul Rahman
 
Classification with ant colony optimization
Classification with ant colony optimizationClassification with ant colony optimization
Classification with ant colony optimizationkamalikanath89
 
Ant colony optimization based routing algorithm in various wireless sensor ne...
Ant colony optimization based routing algorithm in various wireless sensor ne...Ant colony optimization based routing algorithm in various wireless sensor ne...
Ant colony optimization based routing algorithm in various wireless sensor ne...Editor Jacotech
 
Ant Colony Algorithm
Ant Colony AlgorithmAnt Colony Algorithm
Ant Colony Algorithmguest4c60e4
 
Ant Colony Optimization (ACO)
Ant Colony Optimization (ACO)Ant Colony Optimization (ACO)
Ant Colony Optimization (ACO)Mahmoud El-tayeb
 
Ant Colony Optimization for Load Balancing in Cloud
Ant  Colony Optimization for Load Balancing in CloudAnt  Colony Optimization for Load Balancing in Cloud
Ant Colony Optimization for Load Balancing in CloudChanda Korat
 

What's hot (20)

ant colony optimization
ant colony optimizationant colony optimization
ant colony optimization
 
Solving Quadratic Assignment Problems (QAP) using Ant Colony System
Solving Quadratic Assignment Problems (QAP) using Ant Colony SystemSolving Quadratic Assignment Problems (QAP) using Ant Colony System
Solving Quadratic Assignment Problems (QAP) using Ant Colony System
 
Lecture 9 aco
Lecture 9 acoLecture 9 aco
Lecture 9 aco
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimization
 
Ant colony Optimization
Ant colony OptimizationAnt colony Optimization
Ant colony Optimization
 
FAninal aco
FAninal acoFAninal aco
FAninal aco
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimization
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimization
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimization
 
Final project
Final projectFinal project
Final project
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimization
 
Ant Colony Optimization
Ant Colony OptimizationAnt Colony Optimization
Ant Colony Optimization
 
Ant colony optimization
Ant colony optimizationAnt colony optimization
Ant colony optimization
 
ant colony algorithm
ant colony algorithmant colony algorithm
ant colony algorithm
 
Tsp problem
Tsp problemTsp problem
Tsp problem
 
Classification with ant colony optimization
Classification with ant colony optimizationClassification with ant colony optimization
Classification with ant colony optimization
 
Ant colony optimization based routing algorithm in various wireless sensor ne...
Ant colony optimization based routing algorithm in various wireless sensor ne...Ant colony optimization based routing algorithm in various wireless sensor ne...
Ant colony optimization based routing algorithm in various wireless sensor ne...
 
Ant Colony Algorithm
Ant Colony AlgorithmAnt Colony Algorithm
Ant Colony Algorithm
 
Ant Colony Optimization (ACO)
Ant Colony Optimization (ACO)Ant Colony Optimization (ACO)
Ant Colony Optimization (ACO)
 
Ant Colony Optimization for Load Balancing in Cloud
Ant  Colony Optimization for Load Balancing in CloudAnt  Colony Optimization for Load Balancing in Cloud
Ant Colony Optimization for Load Balancing in Cloud
 

Viewers also liked

Genetic Algorithms and Ant Colony Optimisation (lecture slides)
Genetic Algorithms and Ant Colony Optimisation (lecture slides)Genetic Algorithms and Ant Colony Optimisation (lecture slides)
Genetic Algorithms and Ant Colony Optimisation (lecture slides)Dagmar Monett
 
Ant Colony Optimization: Routing
Ant Colony Optimization: RoutingAnt Colony Optimization: Routing
Ant Colony Optimization: RoutingAdrian Wilke
 
EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516
EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516
EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516Yasser Mohammed
 
Intelligent water drops (Persian)
Intelligent water drops (Persian)Intelligent water drops (Persian)
Intelligent water drops (Persian)mortezaT
 
Bee algorithm
Bee algorithmBee algorithm
Bee algorithmkousick
 
Dijkstra’s algorithm
Dijkstra’s algorithmDijkstra’s algorithm
Dijkstra’s algorithmfaisal2204
 
Show ant-colony-optimization-for-solving-the-traveling-salesman-problem
Show ant-colony-optimization-for-solving-the-traveling-salesman-problemShow ant-colony-optimization-for-solving-the-traveling-salesman-problem
Show ant-colony-optimization-for-solving-the-traveling-salesman-problemjayatra
 
Soft computing (ANN and Fuzzy Logic) : Dr. Purnima Pandit
Soft computing (ANN and Fuzzy Logic)  : Dr. Purnima PanditSoft computing (ANN and Fuzzy Logic)  : Dr. Purnima Pandit
Soft computing (ANN and Fuzzy Logic) : Dr. Purnima PanditPurnima Pandit
 
web clustering engines
web clustering enginesweb clustering engines
web clustering enginesArun TR
 
ECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATION
ECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATIONECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATION
ECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATIONMln Phaneendra
 
Economic operation of Power systems by Unit commitment
Economic operation of Power systems by Unit commitment Economic operation of Power systems by Unit commitment
Economic operation of Power systems by Unit commitment Pritesh Priyadarshi
 
Web 3.0 The Semantic Web
Web 3.0 The Semantic WebWeb 3.0 The Semantic Web
Web 3.0 The Semantic WebHatem Mahmoud
 

Viewers also liked (19)

Genetic Algorithms and Ant Colony Optimisation (lecture slides)
Genetic Algorithms and Ant Colony Optimisation (lecture slides)Genetic Algorithms and Ant Colony Optimisation (lecture slides)
Genetic Algorithms and Ant Colony Optimisation (lecture slides)
 
Ant Colony Optimization: Routing
Ant Colony Optimization: RoutingAnt Colony Optimization: Routing
Ant Colony Optimization: Routing
 
EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516
EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516
EXTERNAL - Whitepaper - How 3 Cyber ThreatsTransform Incident Response 081516
 
The future internet web 3.0
The future internet  web 3.0The future internet  web 3.0
The future internet web 3.0
 
Intelligent water drops (Persian)
Intelligent water drops (Persian)Intelligent water drops (Persian)
Intelligent water drops (Persian)
 
Pm imp. book
Pm imp. bookPm imp. book
Pm imp. book
 
Bee algorithm
Bee algorithmBee algorithm
Bee algorithm
 
Lecture 16
Lecture 16Lecture 16
Lecture 16
 
Dijkstra’s algorithm
Dijkstra’s algorithmDijkstra’s algorithm
Dijkstra’s algorithm
 
Show ant-colony-optimization-for-solving-the-traveling-salesman-problem
Show ant-colony-optimization-for-solving-the-traveling-salesman-problemShow ant-colony-optimization-for-solving-the-traveling-salesman-problem
Show ant-colony-optimization-for-solving-the-traveling-salesman-problem
 
Soft computing (ANN and Fuzzy Logic) : Dr. Purnima Pandit
Soft computing (ANN and Fuzzy Logic)  : Dr. Purnima PanditSoft computing (ANN and Fuzzy Logic)  : Dr. Purnima Pandit
Soft computing (ANN and Fuzzy Logic) : Dr. Purnima Pandit
 
Economic load dispatch problem solving using "Cuckoo Search"
Economic load dispatch problem solving using "Cuckoo Search"Economic load dispatch problem solving using "Cuckoo Search"
Economic load dispatch problem solving using "Cuckoo Search"
 
web clustering engines
web clustering enginesweb clustering engines
web clustering engines
 
ECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATION
ECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATIONECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATION
ECONOMIC LOAD DISPATCH USING PARTICLE SWARM OPTIMIZATION
 
Edge detection
Edge detectionEdge detection
Edge detection
 
Economic operation of Power systems by Unit commitment
Economic operation of Power systems by Unit commitment Economic operation of Power systems by Unit commitment
Economic operation of Power systems by Unit commitment
 
Soft computing
Soft computingSoft computing
Soft computing
 
Grey wolf optimizer
Grey wolf optimizerGrey wolf optimizer
Grey wolf optimizer
 
Web 3.0 The Semantic Web
Web 3.0 The Semantic WebWeb 3.0 The Semantic Web
Web 3.0 The Semantic Web
 

Similar to ANT COLONY OPTIMIZATION FOR IMAGE EDGE DETECTION

Ant Colony Optimization algorithms in ADSA
Ant Colony Optimization algorithms in ADSAAnt Colony Optimization algorithms in ADSA
Ant Colony Optimization algorithms in ADSAALIZAIB KHAN
 
antcolonyoptimization-130619020831-phpapp01.pdf
antcolonyoptimization-130619020831-phpapp01.pdfantcolonyoptimization-130619020831-phpapp01.pdf
antcolonyoptimization-130619020831-phpapp01.pdfnrusinhapadhi
 
Heuristic approach optimization
Heuristic  approach optimizationHeuristic  approach optimization
Heuristic approach optimizationAng Sovann
 
Ants coony optimiztion problem in Advance analysis of algorithms
Ants coony optimiztion problem in Advance analysis of algorithmsAnts coony optimiztion problem in Advance analysis of algorithms
Ants coony optimiztion problem in Advance analysis of algorithmsALIZAIB KHAN
 
Introduction to Random Walk
Introduction to Random WalkIntroduction to Random Walk
Introduction to Random WalkShuai Zhang
 
53564379-Ant-Colony-Optimization.ppt
53564379-Ant-Colony-Optimization.ppt53564379-Ant-Colony-Optimization.ppt
53564379-Ant-Colony-Optimization.pptAhmedSalimJAlJawadi
 
Nearest Neighbor Customer Insight
Nearest Neighbor Customer InsightNearest Neighbor Customer Insight
Nearest Neighbor Customer InsightMapR Technologies
 
Modèle de coordination du groupe de robots mobiles
Modèle de coordination du groupe de robots mobilesModèle de coordination du groupe de robots mobiles
Modèle de coordination du groupe de robots mobilesAkrem Hadji
 
Chapter10 image segmentation
Chapter10 image segmentationChapter10 image segmentation
Chapter10 image segmentationasodariyabhavesh
 
Oxford 05-oct-2012
Oxford 05-oct-2012Oxford 05-oct-2012
Oxford 05-oct-2012Ted Dunning
 
Reducing Concentration Uncertainty in Geological Structures by Conditioning o...
Reducing Concentration Uncertainty in Geological Structures by Conditioning o...Reducing Concentration Uncertainty in Geological Structures by Conditioning o...
Reducing Concentration Uncertainty in Geological Structures by Conditioning o...Amro Elfeki
 
Fast Single-pass K-means Clusterting at Oxford
Fast Single-pass K-means Clusterting at Oxford Fast Single-pass K-means Clusterting at Oxford
Fast Single-pass K-means Clusterting at Oxford MapR Technologies
 
Optimization of water distribution systems design parameters using genetic al...
Optimization of water distribution systems design parameters using genetic al...Optimization of water distribution systems design parameters using genetic al...
Optimization of water distribution systems design parameters using genetic al...Tanay Kulkarni
 
Intelligent Handwriting Recognition_MIL_presentation_v3_final
Intelligent Handwriting Recognition_MIL_presentation_v3_finalIntelligent Handwriting Recognition_MIL_presentation_v3_final
Intelligent Handwriting Recognition_MIL_presentation_v3_finalSuhas Pillai
 

Similar to ANT COLONY OPTIMIZATION FOR IMAGE EDGE DETECTION (20)

Ant Colony Optimization algorithms in ADSA
Ant Colony Optimization algorithms in ADSAAnt Colony Optimization algorithms in ADSA
Ant Colony Optimization algorithms in ADSA
 
antcolonyoptimization-130619020831-phpapp01.pdf
antcolonyoptimization-130619020831-phpapp01.pdfantcolonyoptimization-130619020831-phpapp01.pdf
antcolonyoptimization-130619020831-phpapp01.pdf
 
Heuristic approach optimization
Heuristic  approach optimizationHeuristic  approach optimization
Heuristic approach optimization
 
Ants coony optimiztion problem in Advance analysis of algorithms
Ants coony optimiztion problem in Advance analysis of algorithmsAnts coony optimiztion problem in Advance analysis of algorithms
Ants coony optimiztion problem in Advance analysis of algorithms
 
Introduction to Random Walk
Introduction to Random WalkIntroduction to Random Walk
Introduction to Random Walk
 
20191019 sinkhorn
20191019 sinkhorn20191019 sinkhorn
20191019 sinkhorn
 
ACM 2013-02-25
ACM 2013-02-25ACM 2013-02-25
ACM 2013-02-25
 
Optim_methods.pdf
Optim_methods.pdfOptim_methods.pdf
Optim_methods.pdf
 
53564379-Ant-Colony-Optimization.ppt
53564379-Ant-Colony-Optimization.ppt53564379-Ant-Colony-Optimization.ppt
53564379-Ant-Colony-Optimization.ppt
 
Nearest Neighbor Customer Insight
Nearest Neighbor Customer InsightNearest Neighbor Customer Insight
Nearest Neighbor Customer Insight
 
Modèle de coordination du groupe de robots mobiles
Modèle de coordination du groupe de robots mobilesModèle de coordination du groupe de robots mobiles
Modèle de coordination du groupe de robots mobiles
 
Chapter10 image segmentation
Chapter10 image segmentationChapter10 image segmentation
Chapter10 image segmentation
 
Oxford 05-oct-2012
Oxford 05-oct-2012Oxford 05-oct-2012
Oxford 05-oct-2012
 
Clustering - ACM 2013 02-25
Clustering - ACM 2013 02-25Clustering - ACM 2013 02-25
Clustering - ACM 2013 02-25
 
Reducing Concentration Uncertainty in Geological Structures by Conditioning o...
Reducing Concentration Uncertainty in Geological Structures by Conditioning o...Reducing Concentration Uncertainty in Geological Structures by Conditioning o...
Reducing Concentration Uncertainty in Geological Structures by Conditioning o...
 
Fast Single-pass K-means Clusterting at Oxford
Fast Single-pass K-means Clusterting at Oxford Fast Single-pass K-means Clusterting at Oxford
Fast Single-pass K-means Clusterting at Oxford
 
Research Poster_3
Research Poster_3Research Poster_3
Research Poster_3
 
Acoseminar
AcoseminarAcoseminar
Acoseminar
 
Optimization of water distribution systems design parameters using genetic al...
Optimization of water distribution systems design parameters using genetic al...Optimization of water distribution systems design parameters using genetic al...
Optimization of water distribution systems design parameters using genetic al...
 
Intelligent Handwriting Recognition_MIL_presentation_v3_final
Intelligent Handwriting Recognition_MIL_presentation_v3_finalIntelligent Handwriting Recognition_MIL_presentation_v3_final
Intelligent Handwriting Recognition_MIL_presentation_v3_final
 

ANT COLONY OPTIMIZATION FOR IMAGE EDGE DETECTION

  • 1. Guided By : Prof. Vineet Garg. Presented By: Sulabh Pal,Suresh Kumar, Himanshu Srivastava. Date: April 27th 2012.
  • 2.
  • 3. • Section I (Introduction) • Historical Background Sulabh • Ant System Pal • Section II • The proposed scheme Himanshu • Experimental results Srivastava • Section III (Application+Conclusions) • Application • Conclusions and • References Suresh Kumar
  • 4.
  • 6.
  • 7. • Discrete optimization problems difficult to solve • “Soft computing techniques” developed in past ten years: o Genetic algorithms (GAs)  based on natural selection and genetics o Ant Colony Optimization (ACO)  modeling ant colony behavior
  • 8. • Optimization Technique Proposed by Marco Dorigo in the early ’90 • Often applied to TSP (Travelling Salesman Problem): shortest path between n nodes
  • 9.
  • 10.
  • 11. • A stochastic construction procedure • Probabilistically build a solution • Iteratively adding solution components to partial solutions - Heuristic information - Pheromone trail • Reinforcement Learning reminiscence • Modify the problem representation at each iteration
  • 12. • Ants work concurrently and independently • Collective interaction via indirect communication leads to good solutions
  • 13. Basic Phenomenon of Ant For Finding Food
  • 14.
  • 15. Random Way To Find The Way
  • 16. • The pheromone concentration on trail B will increase at a higher rate than on A, and soon the ants on route A will choose • Since the route B is to follow route B shorter, the ants on this • Since most ants will no path will complete the longer travel on route A, and travel more times and since the pheromone is thereby lay more volatile, trail A will start pheromone over it. evaporating • Only the shortest route will remain!
  • 17. • (a) The initial graph with distances. • (b) At time t = 0 there is no trail on the graph edges. • (c) At time t = 1 trail is stronger on shorter edges.
  • 18. • Ants (blind) navigate from nest to food source • Shortest path is discovered via pheromone trails o each ant moves at random o pheromone is deposited on path o ants detect lead ant’s path, inclined to follow o more pheromone on path increases probability of path being followed
  • 19. • Starting node selected at random • Path selected at random o based on amount of “trail” present on possible paths from starting node o higher probability for paths with more “trail” • Ant reaches next node, selects next path • Continues until reaches starting node • Finished “tour” is a solution
  • 20. • A completed tour is analyzed for optimality • “Trail” amount adjusted to favor better solutions o better solutions receive more trail o worse solutions receive less trail o higher probability of ant selecting path that is part of a better-performing tour • New cycle is performed • Repeated until most ants select the same tour on every cycle (convergence to solution)
  • 21. • Algorithm in Pseudocode: o Initialize Trail o Do While (Stopping Criteria Not Satisfied) – Cycle Loop  Do Until (Each Ant Completes a Tour) – Tour Loop  Local Trail Update  End Do  Analyze Tours  Global Trail Update o End Do
  • 22.
  • 23. NATURE COMPUTER SCIENCE NATURAL HABITAT GRAPH(NODES AND EDGES) NEST AND FOOD NODES IN GRAPH;START AND DESTINATION ANTS AGENTS; OUR ARTIFICIAL AGENTS VISIBILITY THE RECIPROCAL OF DISTANCE PHEROMONES ARTIFICIAL PHEROMONES FORAGING BEHAVIOUR Random walk through graph (guided by pheromones)
  • 24. Scheme:  Construct ant solutions  Define attractiveness τ, based on experience from previous solutions  Define specific visibility function, η, for a given problem (e.g. distance)  Ant walk  Initialize ants and nodes (states)  Choose next edge probabilistically according to the attractiveness and visibility  Each ant maintains a tabu list of infeasible transitions for that iteration  Update attractiveness of an edge according to the number of ants that pass  through
  • 25.  Pheromone update  Parameter is called evaporation rate  Pheromones = long-term memory of an ant colony  ρ small low evaporation slow adaptation  ρ large high evaporation fast adaptation  Note: rules are probabilistic, so mistakes can be made!  “new pheromone” or Δτ usually contains the base  attractiveness constant Q and a factor that you want to  optimize  (e.g. ) Q/length of tour
  • 28. Edge The detection proposed ACO Original image Edge image The compensatory edge image The improved result
  • 29. (Initialization) (a). Original image (b). Traditional edge (c). Placing ants on all endpoints detection of (a).
  • 30. 0≦q≦1 If q≦q0 q0=0.9 q=0.5 ≦q0 q=0.3 ≦q0 otherwise 1 2 (Walk) Pixel=1 Pixel=1 Pixel=1 Pixel=1 Pixel=1 Pixel=100 Pixel=100 Pixel=1 Pixel=100 Pixel=100 Pixel=100 Pixel=100 Pixel=1 Pixel=100 Pixel=100 G=24,V=0.2 G=74,V=74 Pixel=100 Pixel=1 Pixel=100 Pixel=100 Pixel=100 Pixel=1 Pixel=100 Pixel=100 Pixel=100 Pixel=100
  • 31. 0≦q≦1 If q≦q0 q0=0.9 q=0.95≧q0 q=0.55≦q0 otherwise (Walk) Pixel=1 Pixel=1 Pixel=1 Pixel=1 Pixel=1 0.02 0.000001 Pixel=100 Pixel=100 Pixel=1 Pixel=100 Pixel=100 0.03 Pixel=100 Pixel=100 Pixel=1 Pixel=100 Pixel=100 High prob Pixel=100 Pixel=1 Pixel=100 Pixel=100 Pixel=100 Pixel=1 Pixel=100 Pixel=100 Pixel=100 Pixel=100
  • 32. (Stop) (a). Two ants meet face to face (b). The ant touches the other’s track. (c). Blind alley (d). Measurement of track density.
  • 33. a. Original image c. 200 iterations e. =0.9, =0.8 g. 200 iterations =0.9, =0.8 b. Sobel edge detector d. Combine b, c f. Combine b, e d. Combine b, g
  • 34. a. Original image b. Canny edge detector d. The improved result from (b) with 160 iterations.
  • 35. a. Original image b. Canny edge detector c. Traditional ACO d. The proposed approach
  • 36.
  • 37. ROUTING IN COMM. NETWORKS Routing task is performed by Routers. Routers use “Routing Tables” to direct the data. If your destination is node 5 next node to 3 4 6 3 1 5 2 2
  • 38. ROUTING IN COMM. NETWORKS Problem statement • Dynamic Routing At any moment the pathway of a message must be as small as possible. (Traffic conditions and the structure of the network are constantly changing) • Load balancing Distribute the changing load over the system and minimize lost calls.
  • 39. Computation time comparison per 20 iterations.
  • 40. Comparision Between ACO and Genetic Algorithm
  • 42. Application • TSP • QAP • Scheduling • VRP • Telecommunication Network • Graph Coloring • Water Distribution Network • Swarm Intelligence • etc
  • 44. URBAN SOLID WASTE COLLECTION TRAFFIC FLOW OPTIMIZATION
  • 45. • Positive Feedback accounts for rapid discovery of good solutions • Distributed computation avoids premature convergence • The greedy heuristic helps find acceptable solution in the early solution in the early stages of the search process. • The collective interaction of a population of agents.
  • 46. • Slower convergence than other Heuristics • Performed poorly for TSP problems larger than 75 cities. • No centralized processor to guide the AS towards good solutions
  • 47.  Can solve certain NP-Hard problems in Polynomial time  Directed-Random Search  Allows a balance between using previous knowledge and exploring new  solutions  Positive feedback for good solutions/Negative feedback for  bad solutions  Approximately convergent  Optimal if not absolutely correct solutions  In certain examples of ACO, no one “ant”  is required to actually complete an accurate  solution
  • 48.  Problem specific  Limited to problems that can be simulated by graphs and optimized  Coding difficulties for different problems  Ineffective utilization of previously acquired information,  specifically the global solution  Depending on the design of the algorithm, it can  converge towards a (less optimal) solution.
  • 49.  We might like to add factors to minimize the time it takes  to reach an acceptable solution.  Use the elements of previous solutions  This allows for faster convergence  As we construct more and more solutions, there is more  information available about the probable “right” choices to make  The decision making process might weigh exploration vs.  heuristic value
  • 50.  Ant System: what we just went over  Ant Colony System:  Pseudo-random proportion rule:  at each decision point for an ant, it has a probability (1-q0) of using the  same probability function as in the Ant System or q0 of picking the best  next node based on previous solutions
  • 51.  Global Trail Update: only the  best solution since the start of  the computation will globally  update its pheromones  Local Trail Update: all ants  consume/decrease  pheromones along the path  that they travel  Elitist Ant System:  Both the global solution and each ant update their edges with  pheromones on each iteration  Applications
  • 52. • The proposed method efficiently improves edge images detected by traditional approaches like Sobel or Canny. • The computation time of the proposed method is shorter than traditional ACO.
  • 53. Dorigo M. and G. Di Caro (1999). The Ant Colony Optimization Meta-Heuristic. In D. Corne, M. Dorigo and F. Glover, editors, New Ideas in Optimization, McGraw-Hill, 11-32. • M. Dorigo and L. M. Gambardella. Ant colonies for the traveling salesman problem. BioSystems, 43:73– 81, 1997. • M. Dorigo and L. M. Gambardella. Ant Colony System: A cooperative learning approach to the traveling salesman problem. IEEE Transactions on Evolutionary Computation, 1(1):53–66, 1997. • G. Di Caro and M. Dorigo. Mobile agents for adaptive routing. In H. El-Rewini, editor, Proceedings of the 31st International Conference on System Sciences (HICSS-31), pages 74–83. IEEE Computer Society Press, Los Alamitos, CA, 1998. • M. Dorigo, V. Maniezzo, and A. Colorni. The Ant System: An autocatalytic optimizing process. Technical Report 91-016 Revised, Dipartimento di Elettronica,Politecnico di Milano, Italy, 1991. • L. M. Gambardella, ` E. D. Taillard, and G. Agazzi. MACS-VRPTW: A multiple ant colony system for vehicle routing problems with time windows. In D. Corne, M. Dorigo, and F. Glover, editors, New Ideas in Optimization, pages 63–76. McGraw Hill, London, UK, 1999. • L. M. Gambardella, ` E. D. Taillard, and M. Dorigo. Ant colonies for the quadratic assignment problem. Journal of the Operational Research Society,50(2):167–176, 1999. • V. Maniezzo and A. Colorni. The Ant System applied to the quadratic assignment problem. IEEE Transactions on Data and Knowledge Engineering, 11(5):769–778, 1999. • Gambardella L. M., E. Taillard and M. Dorigo (1999). Ant Colonies for the Quadratic Assignment Problem. Journal of the Operational Research Society, 50:167-176.
  • 54. D. Martens, M. D. Backer, R. Haesen, J. Vanthienen, M. Snoeck, and B. Baesens, “Classification with ant colony optimization,” IEEE Trans. on Evolutionary Computation, vol. 11, pp. 651–665, Oct. 2007. • R. S. Parpinelli, H. S. Lopes, and A. A. Freitas, “Data mining with an ant colony optimization algorithm,” IEEE Trans. on Evolutionary Computation, vol. 6, pp. 321–332, Aug. 2002. • S. Ouadfel and M. Batouche, “Ant colony system with local search for Markov random field image segmentation,” in Proc. IEEE Int. Conf.on Image Processing, Barcelona, Spain, Sep. 2003, pp. 133–136. • S. L. Hegarat-Mascle, A. Kallel, and X. Descombes, “Ant colony optimization for image regularization based on a nonstationary Markov modeling,” IEEE Trans. on Image Processing, vol. 16, pp. 865–878, Mar. 2007. • A. T. Ghanbarian, E. Kabir, and N. M. Charkari, “Color reduction based on ant colony,” Pattern Recognition Letters, vol. 28, pp. 1383–1390, Sep. 2007. • A. R. Malisia and H. R. Tizhoosh, “Image thresholding using ant colony optimization,” in Proc. Canadian Conf. on Computer and Robot Vision, Quebec, Canada, Jun. 2006, pp. 26–26. • R. C. Gonzalez and R. E. Woods, Digital image processing. Harlow: Prentice Hall, 2007. • H. Nezamabadi-Pour, S. Saryazdi, and E. Rashedi, “Edge detection using ant algorithms,” Soft Computing, vol. 10, pp. 623–628, May 2006. • D.-S. Lu and C.-C. Chen, “Edge detection improvement by ant colony optimization,” Pattern Recognition Letters, vol. 29, pp. 416–425, Mar.2008.
  • 55. Thank you for your attention!