【摘要】 本文把汉字识别归结为无记忆信道对离散信源的信息传输模型。由此出发导出了正确识别率、识别速度的计算公式,分析了影响正确识别率和识别速度的各种因素。给出了正确识别率与被识别字域大小的关系曲线。曲线表明,出现概率越大的汉字对正确识别率的贡献也越大。在汉字综合频度表的6763个汉字中,出现概率大的前4081个汉字对正确识别率的贡献为99.9％,而余下的2682个汉字对正确识别率的贡献仅仅为0.1％。 文中还对提高识别速度的途径进行了探讨,并作了模拟实验,给出了具有启示性的实验结果。更多还原

【Abstract】 In this paper, Chinese Character Recognition is concluded as information carry model of discrete source in zero memory channel. From this, calculating formula of correct recognizable rate and recognizable speed is drawn. At the same time, this paper has analysed factor that makes impact on correct recognizable rate and recognizable speed. Relation’s curve of correct recognizable rate to size of Chinese character’s region is given. It makes clear, Chinese character’s probability is biger and its contribution to correct recognizble rate is biger. The 6763 Chinese characters in synthetical frequency table are sequenced according to probability from big to small. In these, the ahead 4081 Chinese characters contribution is 99.9%. But contribution of the leftovers 2682 characters only is 0.1%.This paper probes into the way of raise recognizable speed. Simulated experiment on computer give in a sense have inspired result.更多还原