使用该B-P神经网络作为汉字的分类器,可以大大提高车牌汉字的识别率。
These features are used to train a B-P neural network, it is a classifier and can improve greatly the recognition rate of Chinese characters.
由于车牌汉字往往存在笔画断裂或粘连、字迹模糊、字形倾斜等问题,导致难以采用常规的方法对其进行识别。
For the reason that it's difficult to recognize license plate Chinese characters through a conventional method because of conjoint or broken strokes, illegible writing and inclined font.
传统的车牌汉字的特征提取方法是在具有先验知识的情况下进行的,先验知识的好坏对结果有着非常重要的影响。
Presently traditional method of car plate Chinese character feature extraction is based on prior knowledge which is important for the result.
应用推荐