Image Interpolation Technique And Its Application Based On PDE

In recent years, the application of PDE and its theory in all the fields of imageprocessing has attracted more and more attentions. As one kind of important processingmethod in the digital image processing field, image interpolation is widely used in imagereconstructing, image zooming, image tone level adjusting, image blurring ,sharp, and soon.In this paper, we firstly analyze some classical algorithms of linear interpolation, onthe basis of which, we propose the PDE interpolation model to get a better e?ect.Secondly, we apply the PDE interpolation model to image segmentation and imagezooming:1. On the basis of partial threshold segmentation,the paper proposes a PDE imagesegmentation algorithm,with which, we can first determine the threshold of each sub-image to get a threshold matrix. Then we use the PDE interpolation model to deal withthe threshold matrix in order to transform it into a new threshold array which has thesame amount of pixels as the original image. After this we can segment the image. Thisalgorithm overcomes the staircase e?ect that is caused by partial threshold segmentation,and then achieves an ideal result.2. The paper also proposes a PDE image zooming algorithm on the basis of linearinterpolation. This algorithm uses PDE interpolation model to process the image zoomedby linear interpolation to get the final zooming image. Moreover, the final zooming imageshows that the PDE interpolation algorithm is better than the traditional interpolationmethods in the aspects of vision and the objective evaluation standard.Finally, to deal with the staircase e?ect caused by TV-norm, we introduce an imagedenoising algorithm based on TV-Stokes equation.