The multi-objective image fast segmentation in complex traffic environment

Abstract

Because of the zoning inadequate of the common two-dimensional histogram and large amount of the two-dimensional Otsu method. In this paper, an improved two-dimensional Otsu method and Quantum Particle Swarm optimization algorithm search for the optimal threshold had been used to multi-objective image segmentation in complex traffic environment. First proposed the Two-dimensional histogram used Filtered gray-scale map-Neighborhood gradient, and then proposed the improved selecting threshold method of the two-dimensional Otsu method. And then, use the improved selecting threshold method as the Quantum Particle Swarm optimization algorithm fitness function to segment image. The results show that, the method presented in this paper can not only get an ideal segmentation results, but also can significantly reduce the computation, achieve fast segmentation.