Research On Human Body Detection Method Of Closed Pipeline Based On Multi-sensor Information Fusion

In the industrial sites such as coal mine,Conveying pipes are widely used to transport materials.Therefore a detection method of the closed conveying pipeline is urgent for detecting human body,stopping operation of the conveying pipeline and sending corresponding alarms under dangerous circumstances.Aiming at the hidden safety hazard of the personnel to fall into the closed pipeline of coal mine underground reversed loader,this paper proposed a human body detection method in the closed pipeline based on the fusion of infrared imaging and capacitance detection information.The main research contents include:This paper summarized application and development trend of existing human detection methods,and compared and analyzed the human body detection methods based on radar,acoustics and infrared imaging.Firstly,it established a human body detection model based on infrared thermal imaging.Aiming at the problem of random blind elements in infrared imaging array,it improved a blind element detection algorithm based on coefficient of variation for infrared sensor array,and used neighborhood mean method for the blind element pixel.The infrared image was processed by image enhancement and threshold segmentation,and the thermal infrared image features of the human body were obtained.Secondly,a variable dielectric constant capacitance human body detection model was established.The relationship between electric field frequency and ambient temperature and dielectric constant was analyzed from the microscopic point of view.A capacitance variation model of the human body mixed into the pipeline was established.Then,the temperature information was extracted by the infrared sensor to compensate the temperature of the capacitor,and the fusion of the multi-source heterogeneous sensor information was completed at the data level.The D-S evidence theory was applied to the decision-level data fusion of heterogeneous multi-sensor information,and the information complement of infrared and capacitive sensors was completed.Finally,a set of human body detection data acquisition system for closed conveying pipeline based on infrared imaging and capacitance detection was built,and the design of modules such as intrinsically safe power supply,infrared thermal image acquisition,capacitance detection and sound and light alarm was completed.The experiment result showed that the interruption caused by environment and other factors would seriously affect human body detection accuracy when human body detection was performed using a single senor;in a human body detection system with heterogeneous multisensors,the detection results of two heterogeneous sensors were complemented with each other,reducing the occurrence possibility of detection result error;the data fusion algorithm of multi sensors compared the basic possibility distribution of two sensors,and the correlation with the basic possibility distribution of two sensors can identify if the human body detection system is abnormal.