基于机器视觉的皮带撕裂图像识别方法

Recognition Method of Belt Tearing Image Based on Machine Vision

为了保障皮带机运行的安全性与连续性,提高皮带运输过程中异常撕裂图像识别的抓图频率和图像清晰度,文章依托机器视觉技术,设计针对皮带撕裂图像的全新识别方法。通过皮带机运行异常画面检测、基于机器视觉的异常画面处理与分析、基于深度学习的皮带撕裂特征提取与分类、皮带机运行撕裂报警,完成对识别方法的设计。通过对比实验,证明此次设计的图像识别方法在实际应用中,可以实现对图像的高频率抓取,并保证识别过程中对图像的处理能够达到更高清晰度水平,为识别效率和识别精度的提升提供有利条件。

In order to ensure the safety and continuity of belt conveyor operation and improve the capture frequency and image definition of abnormal tear image recognition during belt transportation,this paper designs a new recognition method for belt tear image based on machine vision technology.The recognition method is designed by detecting the abnormal image of belt conveyor operation,processing and analyzing the abnormal image based on machine vision,extracting and classifying the characteristics of belt tearing based on deep learning,and alarming the belt conveyor operation tearing.Through comparative experiments,it is proved that the designed image recognition method can achieve high-frequency capture of images in practical applications,and ensure that the image processing in the recognition process can reach a higher definition level,providing favorable conditions for the improvement of recognition efficiency and recognition accuracy.