基于Clifford小波的遥感影像边缘检测

Remote Sensing Image Edge Detection Based on Clifford Wavelet

基于遥感影像结构在方向尺度层面的分解,计算各像素点的Clifford梯度并建立Clifford梯度函数,以此为依据,判断像素是否为边界点,从而得到多幅边缘检测影像。最后将这些边缘影像融合,便可得到最终的边缘检测影像。此方法充分的利用了多光谱影像图层之间的关联性,克服了多维信息难以统一、表达与运算的局限,边缘检测效果优于常用的Sobel与Candy算法。

Based on the decomposition of the remote sensing image structure in terms of scale, phase and direction, the Clifford gradient of each pixel is calculated and the Clifford gradient function is established. Based on this, it is judged whether the pixel is a boundary point, and many edge detection images are obtained. Final y,the edge of the image fusion, you can get the final edge detection images. This method makes full use of the correlation between multi-spectral image layers, overcomes the limitation of multi-dimensional information,the limitation of expression and operation, and the edge detection effect is superior to the commonly used Sobel and Candy algorithm.