The Study On Obtaining Chain Code And Document Layout Analysis

Chain code is a very important tool in digital image processing. It transfers the two-dimensional image to one-dimensional codes. This can greatly reduce the storage space. Each kind of chain code has its own characteristics to facilitate the digital image processing and analysis.Obtaining the chain code is a key step of image processing and analysis. This paper describes several common chain codes and proposes a new way of marking the region contour for tracking. It avoids the possibility of forgetting or repeating to track the contours. The new method can get a variety of chain codes at the same time. This meets the need of the actual application to use different kind of chain codes. Therefore, chain code conversion is not necessary.Region filling is one of the basic topics in the study of graphics. In this paper, a region filling method based on chain code is proposed. It helps to avoid repeating filling the multi-connectivity region and improves the efficiency of filling. On the basis of the study of various polygon scan conversion methods, a new method forwarded in the paper is compared with those methods through the experiments.Document image layout analysis is an important step in optical character recognition (OCR). Based on the analysis of the characteristics in chain codes, the document skew angle is achieved by detecting the rectangles, the lines and words in the document. Characters of rectangle and line are more obvious than those of words. Therefore, detecting rectangles and lines at first is more efficient.Through a lot of experiments, it proved that the method is feasible and effective. And the regional space and paragraph indent are applied to do layout analysis after the correction of angle.