Taking a SEM image of the machined surface as an example, its edge can be extracted by using the local maximum of wavelet s coefficients, and thus the characteristics of a machined surface can be achieved.
图像边缘是图像的重要特征 ,提出了一种基于小波变换的图像多尺度边缘检测算法 ,并以机加工表面SEM图像为对象 ,利用小波系数局部极大值提取其边缘 ,实现机加工表面纹理特征提取。
The sharp variation of gray levels of an image, measured at different scales, can be detected from the local maxima of its wavelet transform.
在不同尺度下图像突变点可以通过它的小波变换局部极大值来检测。
By handling the wavelet transform coefficient modulus in the domain of wavelet transform, and determining the local maxima of the wavelet transform coefficient modulus, it provides information of the image boundaries features.
基于信号与噪声在不同尺度下小波变换系数模不同的变化特征，提出了一种边缘检测方法，该方法通过对图像的小波变换域中由噪声引起的小波变换系数模进行处理，再利用小波变换系数模局部极大值来提取图像的边缘特征，实验结果说明这种特征提取方法可以有效地降低噪声，同时又较准确地提取出图像的边缘。
This paper presents an image edge detection method based on both the algorithm of wavelet local modulus maxim and the fuzzy algorithm on edge detection.
为了更好地对图像边缘进行检测,提出一种基于小波局部极大模和模糊方法相结合的图像边缘检测算法。