基于ICA的有噪图像特征提取研究

Research of noisy image feature extraction based on ICA

讨论了基于ICA的图像去噪方法,给出了基于ICA的图像边缘检测算法,该算法应用于高斯噪声图像,并与传统的边缘提取算法进行了比较。实验结果表明,该算法即使在高水平噪声图像中,也能够提取出比较清晰的图像边缘信息。

This paper discusses the image denoising method based on ICA, and gives the ICA-based image edge detection algorithm. The algorithm is applied to Gaussian noise image, and the traditional edge detection algorithms are compared. Experimental results show that the algorithm, even in high noise image, also able to extract a clearer image edge information.