摘要
针对矿井下皮带运输机无法有效监测大块和锚杆等异物,造成皮带撕裂和皮带跑偏等一系列安全隐患,提出了一种以YOLOv4算法和传统算法相结合的皮带异物安全检测系统。首先,系统为了更快速精准地获取所需图像,在检测区域内对原始图像进行分隔和提取,绘制检测区域,采用小波去噪对图像进行预处理;其次使用以YOLOv4为基础的目标检测算法进行异物监测,运用传统帧差法监测皮带跑偏。系统具有视频录制、异常报警图片保存以及异常报警类型分析的功能,显著提高带式运输机的安全检测效率。
Aiming at a series of safety hazards such as belt tearing and belt misalignment caused by the inability of the mine belt conveyor to effectively monitor foreign objects such as large blocks and bolts,a belt foreign object safety detection system combining YOLOv4 algorithm and traditional algorithm is proposed.Firstly,in order to obtain the required image more quickly and accurately,the system separates and extracts the original image in the detection area,draws the detection area,and uses wavelet denoising to preprocess the image.Secondly,the target detection algorithm based on YOLOv4 is used for foreign object monitoring,and the traditional frame difference method is used to monitor the belt deviation.The system has the functions of video recording,abnormal alarm picture saving and abnormal alarm type analysis,which significantly improves the safety detection efficiency of belt conveyor.
作者
白莹
BAI Ying(Henan Pingbao Coal Industry Co.,Ltd.)
出处
《现代矿业》
CAS
2023年第6期59-62,共4页
Modern Mining