Low-Altitude Airship Remote Sensing Images Matching Based on Improved Hybrid Ant Colony Algorithm

Low-altitude unmanned airship remote sensing is attractive to various applications. However, at present, since the airship is bulky, weak to resist wind, unstable to flight attitude, and can not be equipped with specialized remote sensing sensors, image data processing is confronted with new challenges when traditional data processing methods are used. In this paper, improved hybrid ant colony algorithm (HACA), a new image matching method, is proposed. Firstly we perform a pre-registration process that roughly aligns the image pairs by GPS/electronic compass geolocation. Once the pre-registration is completed, a fine-scale registration process is executed by applying a hybrid algorithm of genetic algorithm (GA) and ant colony algorithm (ACA) based on neighborhood search strategy that is detected by the linear quadtree Morton coding. The image pairs are then matched by using optimal solution obtained from the automatic updates of ant colony pheromone. By compared with traditional genetic algorithm and ant colony algorithm, the improved HACA results show that search calculation time is increased by the maximum 406℅, standard root mean square error of image matching by the best 235℅. The experiment result proves that improved HACA exactly provides an effective method for image matching. The local maxima of the function can be avoided efficiently and search speed of the global optimum is increased greatly.