Wavelet Transform-based Filtering Method To Noise

The noises of the image will affect the comprehension of the image and information extraction from the image. An image is often inevitably corrupted by noises in its acquisition or transmission. The noises in an image degrade severely the following-up image processing tasks, such as image segmentation, coding, feature extraction, and target detection. Therefore, noises reduction has become a very important image pre-processing for improving the quality of image and meeting the needs of processing tasks. How to preserve the details of the high frequency in the image, and to reduce the noises effectively at the same time, is of significance importance. Especially in practice application of reflection digital holography, speckle noises appears inevitable due to the high coherence of laser beam, the speckle noises seriously affects the quality of reconstructed hologram, hinders the extraction of information from the reconstructed image too. The traditional image de-noising methods presented are usually of single algorithm, such as medium filtering in spatial domain, as well as Gauss filtering in frequency domain, and the result of de-noising are not satisfactory. How to eliminate the speckle noise in digital holography has great significance for extraction the information from digital holography and follow-up processing.In this paper, based on the studies and analyzes the characteristics of most common de-noising algorithms, such as the spatial filtering, the Gaussian filtering in the frequency and wavelet transform, and comparing their de-noising effects, Three improved integrated algorithm for de-noising are presented in this paper, The SNR (Signal to Noise Ratio) and the PSNR (Peak signal to Noise Ratio) are used to evaluate the de-noising effect. The simulation experiments demonstrated that the de-noising effect of the integrated algorithm are better than single approach, Finally, the three integrated algorithms are implemented in the de-noising processing of the digital holography reconstructed image, the experiment results proved the better performance of the integrated algorithms than single de-noising algorithm. The integrated algorithms are simple and robust, and have good reference for the research of correlation image processing in other fields.