An Improved Adaptive Ant Colony Algorithm For Intelligent Seamline Detection Of Orthoimage Mosaicking

To automatically search for mosaic seamline of urban large-scale orthophoto, which should avoid buildings, trees and other visually obvious foreground objects, this paper proposes a novel method of searching for mosaic seamline automatically based on an improved adaptive ant colony algorithm. The method based on ant colony algorithm and combined with the thought of Max-Min Ant System, avoids blind search in early phase and premature convergence phenomenon in later period, and improves the ability of global searching for optimal mosaic seamline. Experiment shows that the algorithm can only guarantee the quality of mosaic line, and has a great efficiency, and realizes the large-scale orthophoto intelligent seamless mosaicking.