Studies On Image Super-resolution Reconstruction Algorithms Based On Error-Amended Scheme

Recently, image super-resolution reconstruction is becoming an active research topic in the fields of image processing. Image super-resolution reconstruction can estimate a high-resolution image or a high-resolution image sequence from a single low-resolution image or a low-resolution image sequence having slight difference, correspondingly. Nowadays, this technology is widely used for remote sensing, meteorology, medical imaging, film production, military and so on, playing an important role in social economic development.Image interpolation is one of the most effective methods for image super-resolution reconstruction. Image interpolation can estimate gray values of unknown pixels from adjacent pixels of the original image, and then can generate a new high-resolution image.This paper mainly discusses how to use error-amended scheme to get a super-resolution reconstruction of a single image. First, it introduces the research background and the research significance of super-resolution reconstruction, analyzes the actuality at home and abroad. Then, it elaborates the traditional interpolation methods, such as nearest neighbor interpolation, bilinear interpolation and bicubic interpolation. On the basis of the introduction of error-amended theory, the algorithms are proposed as follows:A novel weighted parabolic interpolation algorithm is proposed. The proposed algorithm incorporates the error-amended part into weighted parabolic interpolation algorithm with Newton interpolation and Lagrange interpolation formula. In order to reduce ringing effects and checkerboard effects, directions of the interpolation points are determined by the Sobel operator, and so the method has an excellent edge-adaptive ability.A novel image zooming algorithm is proposed by using bilinear interpolation and wavelet transformation based on an error-amended sharp edge scheme. The proposed algorithm incorporates the error-amended part into bilinear interpolation. Directions of the interpolation points are determined by the Sobel operator, and then we get a preliminary zoomed image. High-frequency components are got by wavelet transformation, and low-frequency components are replaced by the enhanced amplitude of the original image. A high-resolution image is achieved by inverse wavelet transformation. This algorithm can get clearer and sharper edges due to considerations of global pixels correlation of the original image.A novel image zooming algorithm is proposed by weighted parabolic interpolation and wavelet transformation based on an error-amended sharp edge scheme. The proposed algorithm incorporates the error-amended part into weighted parabolic interpolation algorithm. Directions of the interpolation points are determined by the Sobel operator, and then we get a preliminary zoomed image. High-frequency components are got by wavelet transformation, and low-frequency components are replaced by the enhanced amplitude of the original image. A high-resolution image is achieved by inverse wavelet transformation. This algorithm can get clearer and sharper edges due to considerations of global pixels correlation of the original image.A novel image zooming algorithm is proposed by using improved bilinear interpolation and Contourlet transformation. The proposed algorithm incorporates the error-amended part into bilinear interpolation. Directions of the interpolation points are determined by the Sobel operator, and then we get a preliminary zoomed image. High-frequency components are got by Contourlet transformation, and low-frequency components are replaced by the enhanced amplitude of the original image. A high-resolution image is achieved by inverse Contourlet transformation. This algorithm can get clearer and sharper edges and textures due to considerations of global pixels correlation of the original image.A novel image zooming algorithm is proposed by using improved weighted parabolic interpolation and Contourlet transformation. The proposed algorithm incorporates the error-amended part into weighted parabolic interpolation algorithm. Directions of the interpolation points are determined by the Sobel operator, and then we get a preliminary zoomed image. High-frequency components are got by Contourlet transformation, and low-frequency components are replaced by the enhanced amplitude of the original image. A high-resolution image is achieved by inverse Contourlet transformation. This algorithm can get clearer and sharper edges and textures due to considerations of global pixels correlation of the original image.