Research On The Regular Document Image Fragment Reassembly Based On The Text Of Characteristics

The regular document image fragments reassembly is an important research topic in the field of digital image processing,it is by means of the computer to put a large number of regular document image fragments together. In the judicial, public security and other departments have specialized document destruction tools, the paper documents piled together for cutting destruction. However, many of the destroyed documents again have value, and then it was necessary to restore the fragments. In the face of a large number of regular document fragment, joining together by artificial, the efficiency is very low. Automatic reassembly by using computer,will greatly improve work efficiency. Therefore,how to realize the technology of the regular document fragment reassembly is an important research work.This paper will use the stroke model,image binarization,stroke feature matching,fuzzy recognition, statistics and other methods, proposes a new algorithm based on text strokes features. Through the computer program,the experimental results show that the algorithm is of high efficiency, robustness. The main work and innovation are as follows:1.Introduction of character stroke model. Through the analysis of the text characteristics,establish the stroke model,the pixel lines are abstracted into different strokes,and it is the base of boundary features.2. Stroke contour extraction method. To explore the contour of the stroke pixels in the image boundary, and to judge the type of the stroke.3. Boundary matching algorithm is introduced into matching criteria. Keep matching process to match criterion, reduce errors.