一种基于视觉分析的指针式仪表智能抄读方法

An Intelligent Reading Method of Pointer Instrument Based on Visual Analysis

在电气室设备指针式仪表抄读中,人工抄读易出现漏检、误检和检测不规范以及其他安全问题,故研究高精度的指针式仪表抄读方法来替代传统抄读方法具有重要意义。针对人工抄读和现有指针式仪表检测识别算法出现的诸如误差大、检测流程繁琐等各种问题,设计了一种基于视觉分析的指针式仪表智能抄读方法。该方法通过YOLOv3的特征提取网络对仪表图像进行表盘提取和刻度数字关键点检测,由于提取后的图像可能是通过包含仰视、平视和俯视三种采集姿态在内的不同环境条件拍摄得到的,所以图像会出现一定程度的畸变。为了减少识别误差,还需要进行基于透视变换的倾斜畸变矫正处理,再通过基于霍夫变换概率直线检测和极坐标变换的距离法进行示数判读。多次实验结果表明,该方法在指针式仪表识别的平均准确度达到97.48%,帧速率达到4 fps,并且该方法仍然具有良好的鲁棒性,能够满足实际工程需求。

In the reading of pointer instrument in electrical room, manual reading is prone to missed detection, false detection, non-standard detection and other safety problems. Therefore, it is of great significance to study the high-precision pointer instrument reading method to replace the traditional reading method. Aiming at the problems of manual reading and existing pointer instrument detection and recognition algorithms, such as large error and cumbersome detection process, an intelligent reading method of pointer instrument based on visual analysis is designed. The dial of the instrument image is extracted and the key points of the scale number are detected through the feature extraction network of YOLOv3. Since the extracted image may be taken under different environmental conditions including three acquisition postures of up view, head up view and top view, the image will be distorted to a certain extent. In order to reduce the recognition error, tilt correction processing based on perspective transformation is also required. The distance method based on Hough transform probability line detection and polar coordinate transformation is used for numerical interpretation. The experimental results show that the proposed method in the pointer instrument identification reaches 97.48% and the frame rate reaches 4 fps. Moreover, it still has strong robustness and can meet the actual engineering requirements.