A new fast super-resolution reconstruction method based on non-local means

The high-resolution images are widely used in many fields, such as aerospace, military, medical etc. Some images captured directly sensors can not satisfy the requirements of the practical applications, so they need to be further processed. This paper proposed a new and fast super-resolution reconstruction method based on the non-local means (NLM) for image sequences. The high quality images are obtained by the iterative curvature based interpolation (ICBI). The pixels are classified by adaptive threshold, and the weight matrix is then calculated to form the value of the target point. In the end, the high resolution image is obtained by deconvolution. Experimental results demonstrate that the proposed method is better than that of the original NLM and computational efficient. The proposed method is a good way to obtain a high resolution image.