【作者】 黄穗；

【导师】 顾国庆；

【作者基本信息】 华东师范大学 ， 计算机软件与理论， 2007， 硕士

【摘要】 随着数字图像处理和模式识别技术的不断发展与成熟，图像的编码技术得到了飞速发展，并以其优良的特性在图像处理领域得到越来越多的运用。本文指出了原顶点链编码算法在标定的时候存在的一些问题：在某些情况下会漏标定链编码。本文接着提出一种新的顶点链编码标定方法及其改进方法解决了这个问题。创新之处是通过使用边缘矩阵的结构改进了标定的方法，提高了算法的可靠性，同时又没有牺牲算法的效率。接下来，本文在新的链编码基础上作了一系列应用：嵌套结构分析、填充、区域统计。并且和同类算法相比，有较高的效率。另外，本文又将新的顶点链编码技术应用到围棋棋谱识别领域。在市场上，有较大的需求把传统的物理介质棋谱转换为电子棋谱，但现今却没有成熟且广泛使用的方法。本文首先在分析了棋谱识别问题，用顶点链编码作为主框架，辅以Freeman链编码、投影法等其他模式识别技术来对棋谱进行识别。本文提出的算法，经过实验，比市场上现存的产品在效率上和效果上有较大改进。本文前半部分研究对象是通用算法，后半部分是在棋谱识别这个具体领域的应用。前半部分是后半部分的基础。更多还原

【Abstract】 The technology of coding for image has a rapid growth due to the continual growth and maturation in the field of digital image processing and pattern recognition. And it is used more and more in the field of digital image processing because of its good features.This paper indicates that there are some problems when old vertex chain code labeling algorithm labels the contours of regions. Sometimes old algorithm may miss some of contours. So this paper proposes a new labeling algorithm to solve this problem, and then gives a further solution to optimize this algorithm in efficiency. This new method uses a data structure named edge matrix to improve the old method, enhance the reliability of algorithm, and have good efficiency. And then, this paper also develops a series of applications based on vertex chain code: nested analysis, region filling and region counting. Compared to similar algorithms, these algorithms have better efficiency.This paper also uses new vertex chain code technology to solve the problem of chess manual recognition. There are strong desires to recognize chess manual in paper, but there are no mature method to fulfil it.After analyzing the problem of chess manual recognition, this paper uses new vertex chain code technology as main framework, and also uses freeman chain code technology, projection method and some other algorithms to solve this problem. By amount of experiments, the algorithm proposed by this paper is proved to have obvious enhancement both in quality and efficiency, compared with the similar product in the market.The first half of this paper mainly focuses on common algorithms, and the second half of this paper mainly focuses on specific field (chess manual recognition). The first half is the basis of the second half.更多还原