期刊名称：International Journal of Multimedia and Ubiquitous Engineering
印刷版ISSN：1975-0080
出版年度：2016
卷号：11
期号：11
页码：39
出版社：SERSC
摘要：The Hough Transform algorithm based on RPSO (Reduced Particle Swarm Optimization) is widely used in image detection. But it has several defects, such as being apt to plunge a local extremum and low target detection precision. In order to overcome these defects of the original algorithm, an improved algorithm, which is updated with the mechanism of SA (simulated annealing), is presented in this paper. In the improved algorithm, the output parameters of the Hough Transform was regarded as particle positions, and the opposite value of accumulation array in Hough Transform was employed as a fitness function of RPSO algorithm. Because the mechanism of the SA was involved, the velocities and positions of the particles are updated in real-time in the process of the crossover and Gaussian mutation. Consequently, the ability of converging to global optimum solution is obviously improved. Then, the comparison and analysis of the experiment results between the original algorithm and the improved one have been carried out in the application of the train wheel image detection. The experiment results demonstrate that the accuracy of image detection is evidently increased and the evolution speed is significantly boosted in the proposed algorithm, especially as the image has
关键词：Hough Transform; RPSO (Reduced Particle Swarm Optimization); SA ;(simulated annealing); crossover and Gaussian mutation
