Intelligent recognition method of the typical pointer electric instrument based on machine vision
Wei Hu, Kuang Fan · Proceedings of the 2020 4th International Conference on Electronic Information Technology and Computer Engineering · 2020
In order to solve the problem that existing recognition methods of typical pointer electric instrument which demand high image quality and are greatly influenced by the environment, an intelligent recognition method based on machine vision is proposed. In the process of image processing, an image transformation method is used to eliminate the influence of the shooting Angle. A dynamic threshold segmentation method is used to segment the image, which is not sensitive to the alteration of illumination. By setting different filter parameters, the regions of dashboard, calibration and pointer can be exactly extracted. The subpixel edge extraction method and the circle fitting method are used to extract the features of the dashboard contains radius and rotation center. The method of minimum external rectangle is used to extract the features of calibration and pointer, which contains the information of width, height and rotation angle. The fitted instrument is reconstructed by means of circle fitting and line fitting with the extracted features of dashboard, calibration and pointer. Based on the function between indicator number and the rotation angle of pointer, the indicator number can be calculated. This recognition method can be applied to the indoor and outdoor environment in the substation. Comparing the artificial reading with the indicator number recognized by this recognition method, it is proved that this method provides perfect identification accuracy.