基于ICA阈值优化耦合信息熵的边缘提取算法
An Edge Extraction Algorithm Based on ICA Threshold Optimization and Information Entropy

为了解决传统边缘提取算法对噪声敏感和阈值难以选取，边缘清晰度不高以及边缘不平滑等问题，提出了一种基于ICA阈值优化耦合信息熵的边缘提取算法.首先，基于灰度分布模式将图像分成若干子块，并计算每个子块的分段阈值；然后，为了从大量的分段阈值选择合适的阈值，引入了帝国主义竞争（imperialist competitive algorithm，ICA）优化算法，计算图像的最优阈值，根据获得的最优阈值将每个图像子块划分为不同的均匀区域；最后，通过计算每个均匀区域的信息熵，利用信息熵检测所有处于不同均匀区域的边界像素来提取边缘.实验结果表明：与当前常用的边缘提取算法比较，本文算法具有更高的品质因数与边缘连续性，能够抑制过于微小和琐碎的细节，突出有效的边缘信息，边缘定位精度高且平滑连贯，能够准确地提取目标轮廓.
The traditional edge detection algorithm is characterized by noise sensitivity, difficult threshold selection, low edge sharpness, unsmooth edge and others. In order to solve such problems an edge detection algorithm based on ICA threshold and information entropy is proposed in this paper. Firstly, the image is divided into several sub-blocks based on the gray level distribution, and the segmentation threshold of each sub-block is calculated. Next, in order to select the appropriate threshold from a large number of segmented thresholds, the imperialist competitive optimization algorithm is introduced to calculate the optimal threshold value of the image. Then, each image sub-block is divided into different uniform regions according to the optimal threshold obtained. Finally, the information entropy of each region is calculated, and the edge pixels of all the regions in different uniform regions are detected, using the information entropy. Experimental results show that compared with the commonly used edge detection algorithms, the algorithm described herein has a higher quality factor and edge continuity, is able to suppress the too small and trivial details and highlight the effective edge information. The edge has good positioning accuracy and is smooth and coherent. Therefore, the new algorithm can accurately extract the contour of the target