Design And Implementation On Document Image Recognition System

As computer science and multimedia technology are developing fast, multimedia digital image has become more popular. As a result, digital image processing and image pattern recognition extend to various areas gradually and are being improved and developed continually. Requirement for automation and digital office brought about the transformation from digital document image from paper document. Document image processing and feature extraction from document image has become the focus of current research. This paper introduces the current status and future direction of digital image application, and discusses several methods such as image enhancement, analysis, restoration and compression. It also introduces application of pattern recognition, especially image pattern recognition. The paper describes the development of information recognition of document image, analyzes its limitation, which mainly brought by limitation of optical apparatus itself, and aiming at this deficiency, proposes an effective software solution.Based on analysis of optical mark recognition, a document image processing system based on OMR is discussed, its method and implementation is presented. The key functions of this system according to its processing steps are discussed: image digitalization, image preprocessing, information acquisition and feature extraction of specific information. According to the characteristic of document image, an edge detection algorithm based on special area segmentation in document image is given out. Two different methods: cluster analysis and statistic analysis for feature extraction in the process of mark recognition are also put forward. Based on analysis on the operation performance and correction ratio, we choose the latter as the processing method in this system. At last, presented experiment results and conclusion are given out. Experiment show that, software solution, by the combination of digital image processing and pattern recognition, is able to overcome deficiency of optical apparatus and gain good recognition results.