This document proposes a new type of CAPTCHA called an image arrangement CAPTCHA that uses rearranged image pixels from photographs. It discusses how existing text-based CAPTCHAs can be solved by computers but image recognition remains difficult. The proposed CAPTCHA randomly rearranges pixels from an image into a grid that users must rearrange correctly. Experimental results showed humans could easily solve rearranged image puzzles. The document discusses improving security, accumulating image sources, and addressing usability concerns like response time. It argues image CAPTCHAs could be both secure against computers and entertaining for humans.