改进的阈值图像去噪算法仿真研究

Improved Image Thresholding Denosing Algorithm

研究了图像优化识别问题,图像中噪声经常会影响图像的清晰度,造成图像模糊等。为了更好的去除图像中的噪声,特别是去除图像中细节丰富的区域中的噪声,通常传统的去噪方法难以完成。为了更好的去除图像噪声并较好的保留图像细节信息。在经典的小波软、硬阈值消噪方法的基础上,提出了一种小波包分析的改进方法。小波包变换是一种时频分析的方法,在分析中高频方面优于小波变换,将其应用于图像中噪声的消除。在Matlab上仿真结果表明,此法同时克服了传统阈值方法的缺点,有效提高了图像去除噪声能力,清晰度更高,为图像优化消噪提供了参考。

Image noise often affects the clarity of the images, resulting in image blurring. In order to better eliminate the image noise, especially the noise in the area with rich details, the paper put forward a improved wavelet packet analysis method based on classical wavelet soft and hard threshold denoising method. Wavelet packet transform is a time-frequency analysis method, and the frequency analysis is better than the wavelet transform, therefore it was applied to image noise elimination. Matlab simulation results show that this method overcomes the shortcomings of the traditional threshold method, effectively improves the image, and has higher resolution. The noise elimination effect is better than the wavelet denoising method.