The Research Of Image Super-Resolution Algorithm In Wavelet-Domain Based On Interpolation

Image super-resolution reconstruction is a focus of image processing, recently, it is be widely applied in satellite, remote sensing, military scout, medicinal image and etc. Since image sensors have been used by more and more, people already have got fairly good quality images. However, due to hardware cost and fabrication complexity limitations, how to increase the current resolution level by applying low cost software methods has become a topic of very great interest. So super-resolution is not only important in application, but also significant in theory.It is the super-resolution reconstruction that obtaining high resolution and quality image from a series of low resolution image. This paper mainly studies two algorithm of wavelet domain image super-resolution based on interpolation. because of traditional wavelet transform having two aspect shortcoming, that is, it is not having translation invariability and its limited directional selectivity, however dual-tree complex wavelet transform not only can make up above insufficiency exactly but also has a lot of fine characteristic, stemming from above reasons, text substitutes for traditional wavelet with dual-tree complex wavelet, and introduces image interpolation algorithm based on dual-tree complex wavelet transform. The algorithm enlarges the image by unifying new edge-directed interpolation, bicubic interpolation and dual-tree complex wavelet transform, then, modifies high-frequency coefficient for processing edge. The experiment result indicates that this method is more effective.In order to improve luminance of image, text unifies wavelet fusion and interpolation, suggests image interpolation algorithm based on wavelet fusion. Take advantage of different interpolation method processing the image having the different enlargement effect characteristic, algorithm uses two interpolation method to enlarge one image, then, it fuses two enlarged image. This algorithm has relatively good effect for optimizing image articulation.Lastly, paper concludes the problems of algorithms, and forecasts the work which will be do in future.