Study On Explosive Identification And Location In Blasting Scene Based On Binocular Vision

Through the research and analysis of demolition blasting in the blasting field,it is found that manual loading is still used in the explosive loading.Before the demolition blasting,a large number of manpower is needed to fill the explosives into thousands of blastholes,so the demolition blasting needs to introduce autonomous mobile blasting robots to replace the staff for explosive loading.Therefore,this thesis integrates visual technology into the field of blasting demolition,providing software basis for autonomous mobile explosive loading robot,improving work efficiency and saving cost,so as to promote the process of intelligentization in the field of blasting demolition.In this thesis,the experiment is divided into two stages according to the demolition blasting scene.In the first stage,this thesis conducts the identification experiment of explosives under different environmental factors based on the YOLO convolution neural network,and conducts the ranging and positioning experiment of explosives based on the binocular vision measurement method.By analyzing the experimental results and data,the influence degree and correlation of different environmental factors on the identification and location of explosives are judged,which provides an information basis for the sorting of explosives by autonomous blasting robots.In the second stage,this thesis designs an environmental perception monitoring system and APP.The system can monitor the current environment and working conditions of the mobile blasting robot in real time and make decisions according to the current environmental conditions.At the same time,the APP provides real-time monitoring information to the staff.Finally,the position and size information of the explosive is obtained by measuring the position and size of the explosive,which provides the position and size information of the explosive for the autonomous blasting robot to grasp the explosive.The experimental results show that the technical idea of explosive identification and positioning designed in this thesis is reasonable and feasible,and the explosive identification and positioning system designed in this thesis has certain accuracy and stability.The environmental perception monitoring system designed in this thesis is programmed and implemented on the computer and mobile platform respectively,which has certain theoretical and application value.