急性运动中超分辨率图像重构方法研究

采用当前急性运动中超分辨率图像重构方法得到的重构图像存在全局误差,导致重构图像质量低下,重构效果不佳。为此,提出一种新的急性运动中超分辨率图像重构方法,设计急性运动中超分辨率图像重构模型,将小波稀疏字典作为急性运动中超分辨率图像重构的理论依据。将低分辨率急性运动图像分割成低分辨率图像块,对无噪高分辨率急性运动图像块相应的无噪低分辨率图像块进行分析。通过OMP方法对稀疏系数进行求解,依据得到的稀疏系数估计出高分辨率急性运动图像块的高频小波系数,将高分辨率小波系数急性运动图像块返回高分辨率小波系数急性运动图像,通过逆小波变换得到最终的高分辨率图像,对全局误差进行修正。实验结果表明,采用所提方法得到的重构图像质量高,重构效果好。

Acute movement super-resolution image reconstruction methods

the current acute movement is reconstructed image super-resolution image reconstruction methods have been the global error, lead to low quality of reconstructed images, reconstruct the effect not beautiful.Acute movement for this, put forward a new super-resolution image reconstruction method, the design of acute exercise super-resolution image reconstruction model, on the basis of the establishment of the wavelet sparse dictionary method for wavelet sparse dictionary applied to acute movement of super-resolution image reconstruction theory basis.The low resolution of acute movement into a low resolution image segmentation image block, has no noise and high resolution image block corresponding acute movement without noise low resolution image block is analyzed.Through the OMP method to solve the sparse coefficient, on the basis of the sparse coefficient for estimate the high resolution of acute exercise frequency wavelet coefficients of image block, the high resolution of the wavelet coefficient of acute movement image block is returned high-resolution image wavelet coefficients of acute movement, the final high resolution is obtained by inverse wavelet transform image, the global error correction.Experimental results show that the proposed method of reconstructed image quality is high, the reconstruction effect is good.