新的基于统计熵功率的OCR算法及其DMCU实现  

Novel algorithm and its realization of optical character recognition based on statistical entropy power

作　　者：吴永亮[1] 万旺根[1] 钱锋[1] 徐鸿玮[1] 

机构地区：[1]上海大学通信与信息工程学院,上海200072

出　　处：《计算机工程与应用》2009年第1期195-197,共3页Computer Engineering and Applications

基　　金：国家高技术研究发展计划(863);上海市国际合作项目(No.075107035);上海市教委电路与系统重点学科(No.J50104)~~

摘　　要：使用摄像头进行文字识别最大的问题在于图像抖动。为了有效地消除图像抖动并正确实现文字识别,提出了一种基于统计熵功率的新的识别算法。这种方法将采集到的数据作为随机信号处理。实验证明,此算法计算复杂度低,识别率高,适用于低成本嵌入式系统,在中国台湾俊亿公司24MHZ16位DMCU嵌入式系统上,获得了94%以上的正确识别率。The problem for applying OCR through a common camera is the wobble of the received image data.For the purpose of accurate recognition of Chinese characters,a new algorithm based on the statistical entropy power is proposed in this paper, which transforms the received image data into a set of random signal in order to process.Experimental results have shown that the algorithm has low computation complexity and high recognition rate,making it suitable for low-cost embedded system implementation,and a recognition rate of above 94% has been achieved on the 24 MHZ 16 bit DMCU embedded system which is produced by Taiwan of China chip manufacturer KingBillion.

关 键 词：统计 熵功率 文字识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]