Font Size: a A A

Research On Digital Image Data Compression Algorithm Based On Edge Detection

Posted on:2015-05-22Degree:MasterType:Thesis
Country:ChinaCandidate:L Y JiangFull Text:PDF
GTID:2358330518973206Subject:Education Technology
Abstract/Summary:PDF Full Text Request
With the rapid development of digital multimedia technology and communication technology,digital image has become one of the main carriers of information.The requirements of data storage and transmission speed become higher and higher.The digital image compression technology become one of the focuses in the field of image processing,as it can well solve the problems of data storage,realize the fast transmission on the Internet and real-time processing.Image vectorization is one of the most important aspects in the field of image processing.On the foundation of studying other image compression algorithm and image vectorization algorithm,we present a kind of image compression algorithm based on the image edges.The major contributions of this paper are:1.We introduce three kinds of edge detection operator:the first kind is the classical operator,the second one is based on global optimal,the last one is based on the signal processing,we also give some typical examples of them.In our algorithm,edge information is first detected by Canny operator and thinned by Mathematical Morphology method.The color information is detected according to the geometry information of the image edges.2.In this paper,we introduce four kinds of most commonly used method of curving fitting:the least squares method,the moving least squares method,the Bezier curves,b-spline curves,and make a comprehensive comparison of these methods'properties and characteristics.In our algorithm we choose the b-spline curve fitting which is more local controllable and has smaller amount of calculation to fit the geometry and color information of edges.This is the job of image data compression.3.Raster image is convenient for image display,so image decompression should be rasterized according to the data of vector image edges.In order to prevent the aliasing and ensure the error between raster image and the edge which is shown by the b-spline curve is less than half a pixel,we use midpoint algorithm to rasterize the edge curves.4.According to the raster image edges and the color information of edges,we use the neighborhood averaging method to diffuse the color sources.Experiments show that the algorithm is robust for various images,and only little difference can be found between original image and the corresponding reconstructed raster image by our algorithm.The algorithm can deal with both gray and color images.Besides,on the foundation of the above algorithms,an image vectorization algorithm is implemented in this thesis,and several typical image examples are given as well.The results of these examples indicate the correctness of the image data compression algorithm based on edge information.
Keywords/Search Tags:compression of the image, image vectorization, Canny operator, mathematical morphology method, B-spline fitting, neighborhood averaging
PDF Full Text Request
Related items