基于小波变换的人体图像边缘检测

Human Body Image Edges Detection Based on Wavelet Transform

根据人体图像的特点 ,利用三次样条二进小波 ,采用Mallat快速算法进行小波分解 ,提取了人体图像的正面和侧面边缘 .结果显示 ,提取图像的边缘连续性较好 ,弯曲部分边缘也能准确提取 。

According to the peculiarity of human body image, dyadic wavelet transform is successfully applied to detecting the face and profile edges of human body image and the Mallat algorithm is used in the wavelet decomposition in this paper. The detected edges of human body image have preferable continuity, edges of flexure could also be extracted exactly. And new method has been proved to edges detection of dressed human body image.;