用于图象处理的两种新边界检测算子

通过对线性和非线性拉普拉斯算子的分析，提出了一种新的边界检测算子－中值非线性拉普拉斯算子，该算子无方向性，既可保留边界，又可去除噪声，此外，还提出了另一种用图象灰度变换与边界提取相结合检测边界的方法，由此，产生了另一种非线性边界提取算子－灰度均衡化中值非线性拉普拉斯算子。最后对一幅加了盐－颗粒噪声的人物图象进行了计算机验证。

Two New Operators for Edge Detection on Image Processing

Presents a new kind of operator for edge detection-the median nonlinear Laplacian operator based on the analysis of both linear and nonlinear Laplacian operators.The new operator is nondirective and can maintain the detected edge while removing the noise.Another nonlinear edge extraction operator-the grey flatting dian nonlinear Laplacian(FMNLLAn)operator is also developed bed on the above discussion.The latter operator possesses higher accuracy on edge detection.Both have been proved through experiments.