基于小波变换和曲波变换的图像边缘检测新算法

针对图像处理中的边缘检测问题,提出了一种基于小波变换和曲波变换的图像边缘检测新算法。首先对原始图像进行小波变换得到小波边缘图像;然后对原始图像进行曲波变换并使用Canny算子得到曲波边缘图像;最后基于小波变换的窗口内边缘强度自适应融合算法将小波边缘图像和曲波边缘图像进行融合得到最终边缘图像。该方法结合了小波变换描述图像细节特征的优势和曲波变换处理曲线或直线边缘特征的优势,能全面刻画边缘图像的纹理与细节信息,提高了图像清晰度。仿真实例表明了该算法的有效性。

A novel algorithm of image edge detection based on wavelet transform and curvelet transform

Facing the edge detection problems lying in image processing, a novel  algorithm of image edge detection is put forth based on wavelet transform and curvelet transform. Firstly, we process the original image with the method of wavelet transformation and obtain a wavelet edge image. Then we achieve a curvelet edge image by processing the original image with the method of curvelet transformation and the Canny operator. Finally, the final edge image is acquired by merging the wavelet edge image and the curvelet edge image based for the adaptive fusion algorithm on the edge intensity in wavelet transform window. The algorithm combines the advantage of wavelet transform in describing image detail characteristics and the advantage of curvelet transform in processing curve or straight line edge features. It fully characterizes the details and texture information of the edge image, and improves image clarity. Simulation results verify the effectiveness of the algorithm.