Research On Intelligent Reading Of Pointer Meters In Complex Environment Based On Machine Vision

In the field of monitoring and data acquisition,pointer instruments have the characteristics of sensitivity,high accuracy,strong anti-interference ability,and convenient maintenance.They are widely used in industrial manufacturing sites with strong electromagnetic interference and aerospace fields.At present,many automatic reading algorithms for pointer meters have been proposed,which have a good recognition effect for environments with less interference noise.However,for the interference of the outdoor complex environment,the shadow of the pointer will affect the reading of the meter.Similarly,the reflection of the mirror surface and the water mist on the mirror surface may cause the disappearance of key features such as the scale line.Therefore,how to overcome the interference of the complex outdoor environment is still a puzzling issue.Aiming at the interference of the above-mentioned complex environment,this paper proposes an automatic meter reading method based on random sampling statistical circle detection-similarity rotating pointer template matching,which automatically reads the indicator of the pointer meter in the outdoor complex environment.The main work of this paper is as follows:(1)As the amount of light in the outdoor environment changes all the time,the collected pointer meter images may be blurred and have poor contrast,causing misrecognition of meter readings.For the interference caused by the poor image acquisition quality,this paper uses the BM3 D image filtering denoising method to remove noise interference,and uses the MSRCR image enhancement algorithm based on the dark channel prior to preprocess the acquired images to reduce the impact of complex outdoor lighting.(2)Aiming at the problem of the Hough circle transformation method in detecting arc breaks and the low accuracy of shooting angle offset recognition.this paper proposes a random sampling statistical circle detection method,which uses the optimal density circle area method,Not only solves the problem of the loss of scale arc characteristics in complex environments,but also greatly reduces the time-consuming detection,and at the same time overcomes the interference caused by mirror reflections under complex lighting.(3)In view of the fact that the traditional Hough line detection method is easily affected by the shape of the pointer,while the silhouette method is easily affected by the complex background of the instrument panel in an outdoor environment,this paper proposes a similarity rotation pointer template matching method.It highlights the meter pointer area by calculating the similarity,and solves the problem of wide and narrow pointers by rotating the pointer template matching method,On this basis,it leads to a pointer detection method based on the summation of radial grayscale images,which greatly reduces the time-consuming system recognition.(4)Developed a set of pointer meter reading recognition system for outdoor complex environment.A series of screening parameters are designed through experiments,which can identify multiple types of pointer meters robustly,and an experiment is designed to be analogous to other pointer meter readings.The recognition effect and overall performance of the system are verified.