机构：
商洛学院数学与计算机应用学院;北京工业大学信息与通信工程学院;
中文关键词：
图像分割;;Tsallis相对熵;;高斯分布;;风驱动优化;;粒子群
中文摘要：
本文提出了一种风驱动优化Tsallis相对熵的图像多阈值分割算法。首先分析了Tsallis相对熵阈值分割原理,并将其推广到多阈值分割。利用高斯分布拟合分割后的图像直方图信息,利用Tsallis相对熵作为衡量最佳分割阈值的度量函数。将风驱动优化算法与Tsallis相对熵度量函数结合,求解Tsallis相对熵函数的最优解,提高阈值分割算法的速度。最后将所提算法与穷举法、粒子群算法做比较,并且与经典的Otsu算法和基于二维熵的多阈值分割法进行对比。实验结果表明所提算法速度快、准确性高能够用于图像的多阈值分割。
英文篇名：
Fast Image Segmentation with Multilevel Threshold Based on Tsallis Relative Entropy and Wind-Driven Optimization Algorithm
英文作者：
LI Fenhong;LU Jing;ZHANG Zhiguang;College of Mathematics and Computer Application, Shangluo University;School of Information and Communication Engineering, Beijing University of Technology;
英文摘要：
This paper proposes a fast image-segmentation algorithm with a multilevel threshold based on the Tsallis relative entropy and wind-driven optimization algorithm. First, the principle of the Tsallis relative entropy is analyzed, and single threshold segmentation is extended to multilevel threshold segmentation. Then, a Gauss distribution is used to fit the image histogram information after segmentation, and the Tsallis relative entropy is used to determine the best segmentation threshold. To improve the speed of the threshold-segmentation algorithm, a wind-driven optimization algorithm is used to find the optimal solution of the Tsallis relative-entropy function. Finally, the proposed algorithm is compared with exhaustive and particle swarm optimization algorithms. The proposed algorithm is also compared with the Otsu algorithm and the multi threshold-segmentation method based on two-dimensional entropy. The experimental results show that the proposed algorithm can be used for multi-threshold segmentation of images with high speed and high accuracy.
英文关键词：
image segmentation;;Tsallis relative entropy;;Gaussian distribution;;wind driven optimization;;particle swarm optimization
