Research On An Subdivision Image Interpolation Method On Android Platform

Image interpolation is one of the most basic content in the field of image processing,which main purpose is to obtain unknown pixel values on high resolution images by known pixel values on low resolution images.At present,image interpolation technology involves various aspects of people's lives,such as aerospace,biological engineering,machine vision and so on.With the development of the current mobile Internet,people obtain images through mobile devices mostly.As the mobile operating system Android with the largest number of users in the world,how to process image information efficiently on its platform is more valuable for research and application.This paper aims to study the image interpolation method that can zoom in on the Android platform and maintain edge information efficiently.According to the analysis of existing algorithms and the development requirements of high-definition digital images on the Android platform,this paper proposes an image magnification method based on curve tessellation.The main research contents are as follows:1.In view of the image blur caused by continuity in the image interpolation process,this paper proposes a discontinuous B-spline interpolation based on the Android platform.The algorithm interpolates the image with the traditional 3 times non-uniform B-spline in the smooth region,and sets the appropriate heavy node at the edge pixel to generate the discontinuous B-spline interpolation,so as to generate the same size as the edge pixel value on the edge.Color values to ensure the clarity of the edges of the image.Experiments show that this method can preserve the texture direction of the image well and get better interpolation results.2.Because the discontinuous B-spline interpolation algorithm only considers the image edges in the horizontal and vertical directions of the image,this paper further proposes an interpolation algorithm based on rational Loop subdivision surfaces.The algorithm first converts the image data into a triangular mesh,and then sets reasonable weight coefficients to control the proportion of different pixels on the triangular mesh,which ensures the sharper image edges in the interpolation process and improves the image resolution rate.Since the method is based on triangulation mesh,in addition to processing the edge information in the vertical and horizontal directions of the image,the edge information in the 45° and 135° directions of the image can be better maintained.The two image interpolation algorithms proposed in this paper are all implemented on the Android platform,which improves the image quality and satisfies the Android user's need for fast and efficient high-resolution images.