Research On Image Interpolation And De-Gaussian-Noise Algorithm

As the development of society and the advance in technology, our lives are increasingly colorful. Getting more and more accurate image information is very important. In modern life, Image Processing is used in many fields, such as the biomedical, remote sensing, military, and the many video multimedia and so on. Image Processing include image edge extraction, image segmentation, image enhancement, image restoration, image data compression, etc. The research in this paper focus on the image interpolation and image de-noise.Image interpolation plays an important role during the maintaining of the image quality. The traditional interpolation algorithms, as nearest neighbor interpolation, bilinear interpolation, cubic convolution interpolation, and so on, have a widely application in the early time because of their simple principle, fast operational speed, but the edge of the image is often blurred, resulting in poor image quality. Edge-directed interpolation can be used to change the situation. According to the edge-directed interpolations which already exist, this paper presents a partial repeat edge adaptive interpolation algorithm. Based on the edge of the image and non-edge region using different interpolation function, by reduction of the error during the interpolation process, this algorithm can have a better protection for the image edge features.Noise in the image has great impact to the image analysis and computer vision. Therefore image de-noise is a very important study in the image processing. The image de-Gaussian-noise is a difficult problem, and people have always been looking for effective methods. Traditional Gaussian noise reduction methods use means filtering. Although the calculation of this method is relatively simple, but the image is blurred, the distinction contrast is reduced and the details are lost. The effect of traditional methods is not very good. The wavelet transform in application mathematics has a rapid development; the de-Gaussian-noise methods based on wavelet transform is proposed and achieved a good result. According to the de-Gaussian-noise algorithms which already exist, by combination of shrinking wavelet threshold and Wiener filter, this paper proposed a de-Gaussian-noise algorithm. This algorithm can remove Gaussian noise effectively; increase the Peak Signal-to-Noise Ration; improve the visual characteristics, and it has validity and feasibility in engineering.