基于Renyi熵与PSO算法的图像多级阈值分割

摘要

In image segmentation, Renyi-based thresholding method has obtained widely application because of its remarkable effectiveness. In order to develop the latent ability of Renyi-based method in image segmentation, the method was extended to multi-thresholding field. However, due to the time complexity, the Renyi entropy-based method was very difficult extended to multi-thresholding scenario straightly. To overcome this problem, a fast multi-thresholding method combined with the particle swarm optimization algorithm for complex image segmentation was proposed based on Renyi entropy. The experimental results show that the proposed method can reduce the computation time greatly, and obtain ideal segmentation result.%在图像阈值分割方法中， Renyi熵法因其显著效能而得到大量应用。为了更好地发挥Renyi熵在图像分割中的应用，提出把Renyi熵法扩展到图像多级阈值化问题。然而，由于计算时间复杂度上的高要求，很难把这种有效的技术推广到复杂图像多级阈值化问题。为减少本方法的计算时间，应用粒子群优化算法实施最佳阈值的搜索。实验结果表明，本方法能有效地对图像进行多级分割，并且显著降低计算时间。