小波变换的图像增强
Image Enhancement by Wavelet Transform

本文对于水下图拍摄照片不清晰和雾天图受天气影响的问题，提出了一种基于直方图均衡和小波变换相结合的图像增强方法。首先将图像归一化转为灰度图像，再利用二维离散小波变换的方法把图片分为四个子图，接着对低频部分的子图进行直方图均衡的图像增强，并且对其他部分子图进行置零处理，最后将处理过后的四个子图进行二维离散小波逆变换，重构出增强图像。文章进一步对小波变换中的滤波器进行多维度处理，并且对小波变换分解和重构的层次进行进一步分析，结果发现随着滤波器维数和小波变换层数的增加，图像增强的效果更好。
In this paper, an image enhancement method based on histogram equalization and wavelet trans-form is proposed to solve the problems of unclear underwater images and weather affected fog im-ages. First, the image is normalized into gray level image, and then the image is divided into four subgraphs by two-dimensional discrete wavelet transform. Then, the low-frequency subgraphs are enhanced by histogram equalization, and the other subgraphs are zeroized. Finally, the four sub-graphs after processing are reconstructed by two-dimensional discrete wavelet inversion. In this paper, the filter in wavelet transform is processed in multiple dimensions, and the levels of decomposition and reconstruction of wavelet transform are further analyzed. The results show that with the increase of filter dimension and wavelet transform layer, the image enhancement effect is better.