Taeg Sang Cho and colleagues at Massachusetts Institute of Technology, have developed a software, that can cope with any image, including photographs of outdoor scenes. "They diced up 5MB pictures into 400 squares and fed them into the software. Computer then analyses "predominant colours" to try to work out what kind of image is hidden in the jumble. It then refers a image database to roughly arrange "pieces in their likely positions". Having guessed rough arrangement for the pieces, software then examines "pixel colour values" along "boundaries of each piece" and finds those on other nearby pieces that match the closest. It then takes a best guess at the likely neighbouring piece: the computer completed the 400-piece puzzle in just 3 minutes," writes NewScientist. (emphasis added)
Taeg's research paper includes: A content-aware image prior, A probabilistic image jigsaw puzzle solver, and Motion blur removal with orthogonal parabolic exposures.