摘要
对于灰度图像和使用越来越多的彩色图像,已有的算法中存在着边缘检测效果不理想、计算复杂以及对光照或阴影的敏感等问题.针对于此,提出用主成分分析算法对灰度图像进行边缘提取;对于彩色图像,提出先用独立成分分析算法得出一幅具有全局特性的灰度图,再利用主要成分分析算法进行边缘特征提取.检测实验结果显示边缘特征清晰,计算简单,且不受光照和阴影的影响.
Image edge feature extraction plays an important role in image analysis and recognition. For gray image and color image ever-increasing in use, there have been some edge feature extraction methods. There were, however, some problems with them that extraction result was complicated and the image edge was sensitive to the illumination and shadow. Therefore, a new gray edge detection method was proposed base on principal component analysis（PCA） algorithm. For the color image, an independent component analysis（ICA） algorithm was to be used first to analyze and get a serviceable gray image, then the PCA was used to detect gray image edge. Experimental results showed that the edge was clear, free from the influence of illumination and shadow, and the way was simple.
出处
《兰州理工大学学报》
CAS
北大核心
2007年第3期100-103,共4页
Journal of Lanzhou University of Technology
基金
教育部"春晖"计划科研启动资金(02022)
关键词
主成分分析
独立成分分析
边缘检测
principal component analysis
independent component analysis
edge feature extraction
