基于深度图像的带式输送机煤量检测方法

Coal quantity detection of belt conveyor based on depth image

针对现有的带式输送机煤量检测方法会受到井下昏暗环境的影响,识别精度不高的问题,提出一种适用于井下环境的带式输送机煤量检测方法。基于深度图像的获取不受井下昏暗环境影响的特点,以深度相机获取的不同煤量深度图像为研究对象,对其进行滤波处理以滤除干扰信息并增强特征信息,提出一种DID-CNN识别网络对滤波后的煤量深度图像进行特征提取,并最终将煤量分为3个不同类别作为检测结果,该结果可用于胶带机带速的分级调控。结果表明:所提出的煤量检测模型的准确率达到99.3%,模型的F1分数为0.991,平均检测每张图片的时间为0.0243 s。基于深度图像的带式输送机煤量检测方法可以有效消除井下昏暗环境对煤量检测造成的干扰,具有较高的检测精度和较快的处理速度。该方法可为提高带式输送机运输效率、实现节能降耗以及延长设备使用寿命等方面提供支持。

The existing coal quantity detection method of belt conveyor is affected by the underground dark environment resulting in a comparatively lower recognition accuracy and a new one suitable for underground environment is proposed.Based on the fact that the acquisition of depth image is not affected by the dark underground environment,the depth images of different coal amounts obtained by the depth camera are taken as the research object,and the filtering processing is carried out to filter out the interference information and enhance the feature information.A DID-CNN recognition network is determined to extract the features of the filtered coal depth image,and the coal amount finally is divided into three different categories as the detection results,which can be used for the grading control of the belt speed of the belt conveyor.The results show that the accuracy of the proposed coal quantity detection model is 99.3%,the F1 score of the model is 0.991,and the average detection time of each image is 0.0243 seconds.The research indicate that the coal quantity detection method of belt conveyor based on depth image can effectively eliminate the interference caused by the underground dark environment on the coal quantity detection,with detection accuracy higher and processing faster.This method can provide support for improving the transportation efficiency of belt conveyor,realizing energy saving and consumption reduction,and prolonging the service life of equipment.