The Characters-Extraction Under Complex Color Background

With the rise of digital libraries and the development of multimedia information retrieve and Internet technology, especially, with the use of colours widely, extracting the characters from color image has been attached importance to more and more, and applied to the information management, such as, Web search, automatic processing of complex documents and various image database retrieve. The text embedded in color image has significant meaning, which usually forms brief description or explanation to the relative image. Recognizing and analyzing these texts' regions which are automatically located and extract information of these texts' charaters from image are very helpful for understanding, indexing and retrieving image.To the characters in color image, the arrangement of them is sequential and the color of the font is basically same. The most importance is there are distinct edge-contours between the characters and the background. So, when exacting these characters from color images, it is very important to use the above-mentioned specialities of these characters.So through the analysis of the general algorithm of color image filtering, based on the classical vector median filter, a modified and effective algorithm is presented. This algorithm integrates linear mean filters and vector median filters. It has good restrain characteristics for single pixel noises and single-pixel-width striation noises. It can retain the edge information of color images and do not output new colors in the images after filtering.Through the analysis of color distance and sobel operator, based on them, an edge detection method is presented. And have got an effective result by using this method in edge extraction experiments of processed color image.Aiming at complex color images with sharply different illumination, motion-blur and color diffusion, this paper presents a new method to segment target and background quickly, and get binary images with high quality.After intensive description of the thesises, the system based on a simplification model is investigated. It includes the detail processing: image preprocess, color edge detection, threshold calculation, binary processing, character segment, and contour extraction. Experiments verify the validation of the thesises present in the paper.