Image Super-resolution Reconstruction Based On Wavelet And PDE

The resolution of the image is always an important standard of measuring image qualitystand or fall. The resolution of the image is higher, the detail is richer and the information ismore. So getting high resolution image is the goal of pursuit for people. However, the imageresolution is enhanced through the methods of improving the density of sensor of acquisitiondevice. The cost is high, difficulty is high and technology also has close to capacity. Therefore,an effective method of improving the resolution of the image is the reconstruction techniqueof high resolution image through obtaining the low resolution image. Becausesuper-resolution reconstruction technology does not involve the hardware and the cost is low,The technology have an important prospect in military, medical, industrial, public security,transportation, civil and so on.First, various super-resolution reconstruction methods are introduced. The advantagesand disadvantages of all kinds of algorithm is illustrated and analyzed in detail. Then, therelated theory of knowledge and the application in the image super resolution reconstructionof wavelet method, variational and partial differential equation (PDE) are deeply introduced.Secondly,this section starts from image zooming model of PDE diffusionpost-processing. Several common diffusion models are introduced. And their advantagesand disadvantages are analyzed. At the same time, a minimal surface functional model ispresented. In order to consider the realize reconstruction effect both image regional and edgearea in the image zooming. An improved four order partial differential equation diffusionmodel is proposed, compared with other partial differential equations diffusion model. Theimproved algorithm can fully play an advantage of LLT model and preserve image edge. Atthe same time, minimal surface is leaded in as control function, which can get goodreconstruction effect in flat area.Finally, variation and partial differential equations are analysied from anotherperspective. Wavelet and improved PDE interpolation of image super-resolutionreconstruction algorithm is proposed. In view of the traditional partial differential equationalgorithm can not accurately estimate the image smooth edge area，however lead to false edgeand image blurring. This paper presenting data weighted fitting PDE interpolation algorithm is proposed, which can achieve better positioning for image edge. Aiming at defect of thetraditional wavelet reconstruction algorithm, the enhanced amplitude two times of originalimage is as the low-frequency. And high-frequency components which are got by wavelettransformation are adjusted their coefficient. At last a high resolution image is achieved bywavelet transformation. The experimental results show that this method can give full play tothe advantage of the two algorithms. Not only improves picture resolution and effectivelykeeps the detail information of the original image, but also improves the definition andbrightness of zoomed image.