基于粒子群优化算法和模糊熵的多级阈值图像分割算法

针对现有阈值分割算法利用穷举搜索寻找最优阈值而造成的计算成本较大的问题，提出了一种基于粒子群优化算法和模糊熵的多级阈值图像分割算法。图像分割是图像分析中非常重要的预处理步骤，在提出的方法中，首先选择香农熵和模糊熵作为优化技术的目标函数；然后建立一种基于粒子群优化算法的多层次图像阈值分割，通过最大化香农熵或模糊熵进行图像分割。最后从图像分割数据库中选取Lena、baboon和airplane作为测试图像进行性能分析（包括鲁棒性、效率和收敛性），并与现有的几种阈值分割算法进行比较。结果显示，提出的算法得到了更高PSNR值和更少的分类误差，证明了该算法是一种高效的多级阈值图像分割算法。

Multi-level threshold image segmentation algorithm based on particle swarm optimization and fuzzy entropy

Aiming at the problem that the existing threshold segmentation algorithm used the exhaustive search to find the optimal threshold and the calculation cost is relatively large, this paper came up with a multi-level threshold image segmentation algorithm based on particle swarm optimization and fuzzy entropy. Image segmentation was a very important preprocessing step in image analysis. In the proposed method, it first selected Shannon entropy and fuzzy entropy as the objective function of the optimization technique. Then it established a multi-level image threshold segmentation based on particle swarm optimization algorithm, and performed image segmentation by maximizing Shannon entropy or fuzzy entropy. Finally, it selected Lena, baboon and airplane from the image segmentation database as test images for performance analysis(including robustness, efficiency and convergence), and compared with several existing threshold segmentation algorithms. The results show that the proposed algorithm obtains higher PSNR value and less classification error, which proves that this algorithm is an efficient multi-level threshold image segmentation algorithm.