摘要
为了自动快速地选取适合不同人脸的边缘检测方法的目的.采用了计算原始人脸图像与其边缘检测后所得图像的均方误差和峰值信噪比的方法,做了比较Roberts,Sobel,Prewitt,Canny,Log边缘检测方法的Matlab仿真实验,获得了上述5种边缘检测方法的均方误差、峰值信噪比和检测时间的比对图,得出了不同人脸在不同环境的条件下自动选取合适的边缘检测方法的结论,具有高效、快捷、便利的特点。
In order to automatically select edge detection methods suitable for different faces, a method to calculate the mean-square error and peak signal-to-noise ratio between original image and edge detected image is adopted. Matlab simulation experiments were conducted to compare Roberts, Sobel, Prewitt, Canny and Log. The comparison charts of the mean-square er- ror, peak signal-to-noise ratio and detection time of the five edge detection methods mentioned above were obtained. A conclu- sion that automatically select the appropriate edge detection method for different faces in different environmental conditions was achieved. It＇s very efficient, fast and convenient.
出处
《现代电子技术》
2013年第6期89-92,共4页
Modern Electronics Technique
基金
陕西省科技统筹创新工程计划项目(2011KTCQ01-23)
西安市科技计划项目(CX1129(n))
关键词
边缘检测
微分算子
均方误差
峰值信噪比
edge detection
differential operator
mean-square error
peak signal-to-noise ratio
