Wavelet Transform In The Digital Image Noise Reduction Processing Applications

Digital image processing is developing rapidly; it has been widely used in various fields such as medicine, astronomy, industry examination and remote sensing. Image denoising is a commonly used technology in image processing. Classical method of image denoising is low pass filtering in essence. But low pass filter will obscure the edge information of images when it diminishes the noise effectually. Thus, because wavelet transform has good spatial and frequency characteristics of analyzing, it provides a powerful tool for signal denoising, and overcome the limitations while processing non-steady signal by using traditional methods.Firstly, this paper introduces some basic concepts of digital image processing, describes various important noises and some traditional filters, and gives examples to analyze the effects of traditional filters. Secondly, it elaborates some important theories of wavelet transform, and gives examples to analyze its uses in image denoising. At the same time, puts forward some knowledge and application of edge detection.Lastly, when we process the images which have Gaussian noises interspersed with salt and pepper noise, we discover that the method of combining wavelet transform with adaptive median filtering overcome the defects of which the wavelet threshold denoising is not effective in diminishing the noise of salt and pepper while the general median filtering is also unable to reduce the Gaussian noise. Experiments confirm that the method can diminish the noises while preserve the details of the original image. The result is better than that of traditional filtering, and even better than those of the method by combining wavelet transform with general median filtering or edge detection. After denoising, the distortion of images by the method is very small, and we arrive at the purpose of denoising.