摘要
根据待识别图像中多个仪表的特点,运用边缘检测技术、数学形态学方法、连通域标记法等来定位出多个仪表待识别字符串区域,并运用BP神经网络来识别。经过试验,能准确定位出各个仪表字符串区域,其识别率也达到了95%。
According to the features of multi-instrument regions in the images to be recognized, we have located the multi-string regions with the methods of edge checking, mathematic morphology and connected-through field mark and employed BP NN to recognize them. By experimentation, the result shows that each string region can be accurately located and that the recognition rate has reached 95%.
出处
《苏州科技学院学报（自然科学版）》
CAS
2005年第4期42-45,49,共5页
Journal of Suzhou University of Science and Technology （Natural Science Edition）
