论文部分内容阅读
为对银行支票进行自动分析、验证、检索，该文设计了一种支票自动处理系统，介绍了图像处理及识别技术在该系统中的应用，描述了一种支票要素自动分割方法及针对支票图像的处理与识别算法，成功地实现了强噪声情况下支票部分要素的图像分割和识别，具有较好的稳定性和很强的抗噪声能力。对５５０张真实使用的支票做版面分割、角码区域分割识别实验，其中版面分割正确率为１００％，角码分割识别准确率为９９１％。另外该文实现了一种小波编码压缩算法，在较高的压缩率下获得了较好的图像质量，极大地节约了支票图像的备份空间。
In order to analyze, validate and retrieve bank check automatically, this paper designs an automated check processing system, introduces the application of image processing and recognition technology in this system, and describes a method of automatic segmentation of check element and a new method of checking check images Processing and recognition algorithm, the successful implementation of a strong part of the noise check the image segmentation and identification of elements, with good stability and strong anti-noise ability. 550 pieces of checks were actually used for layout segmentation, corner code area segmentation identification experiment, in which the layout segmentation accuracy rate of 100%, angle code segmentation recognition accuracy of 99.1%. In addition, this paper implements a wavelet coding compression algorithm, which achieves better image quality at higher compression ratio and greatly saves backup space of check images.