【摘要】 提出了一种基于模糊理论的仪表数字快速识别方法。确定的典型特征具有较高的区分度,且计算简单,无需进行数字细线化和大小归一化处理。构造了一种基于模糊识别最大隶属原则的数字识别器,并且采用BP神经网络解决了确定最优权矩阵的难点问题。试验表明:对于常规测量环境中的数字仪表,该方法的识别率高达99%,对7位数字的识别时间不超过30ms,达到了仪表数字识别的速度和准确率要求。更多还原

【Abstract】 Proposed a fast meter digits recognition method based on fuzzy theory.A group of features with high classification capacity is chosen.These features can be easily extracted with no necessity for thinning and size normalization.A fuzzy digit recognizer is constructed based on the maximum membership principle of fuzzy theory.And BP neural network is introduced to assign the weight value of each feature.The results show that for digital meters working in normal measurement environment the recognition rate is higher than 99% and the recognition time of 7 digits is less than 30 ms,which tells that the method is much more applicable for the requirements of the recognition speed and accuracy of digital meters.更多还原