典型图像边缘检测算子的比较与分析

Comparison and Analysis on Typical Image Edge Detection Operators

通过MATLAB编程,采用Roberts算子、Prewitt算子、Sobel算子、Kirsch算子、Log算子和Canny算子等对数字图像做了边缘检测.通过图像的直方图,直观展示了各算子检测图像之间及其与原图的差别.引入高斯噪声和椒盐噪声,并评价了每种算子抑制噪声的能力.这些研究结果对边缘检测算子的选择及改进具有一定的指导意义.

Using MATLAB programming,six edge detection operators were used to detect digital images.The difference between the image detected by each operator and the original image was intuitively demonstrated from the histogram of the image.Gaussian noise and salt-and-pepper noise were introduced to evaluate the anti-noise ability of each operator detection method.These results will provide some reference for the selection and improvement of detection operators in image-edge detections.