Edge Detection of Laser Range Image Based on a Fast Adaptive Ant Colony Algorithm

Abstract

AbstractLaser range imaging is the current priority research areas of airborne lidar. And realizing accurate edge detection of laser range image is the key of completing the subsequent three-dimensional reconstruction. Based on the characteristics of laser range image and the deficiencies of traditional edge detection methods, a new improved fast adaptive ant colony algorithm for edge detection of laser range image has been proposed in this paper. Due to the initial cluster center and the heuristic guiding function used in the algorithm, the randomness and blindness of ants walking are eliminated thoroughly, and the speed of image edge detection is also greatly increased. Meanwhile, thanks to the applied ants’ selection mechanism and updating mechanism varying in contents, the error detection rate and omission factor of edge points as well as noise interference are all avoided, and the accuracy and adaptability of laser range image edge detection are greatly improved as well. Experimental results have shown that, this algorithm is more effective than other edge detection methods, and can meet the requirements of three-dimensional reconstruction.KeywordsAnt Colony AlgorithmEdge DetectionLaser Range ImageThree-Dimensional ReconstructionContrast Experiment