| External force damage is one of the main causes of transmission line safety accidents.In the traditional line patrol mode and video monitoring system,the warning of external damage is not timely enough,and the consumption of human and material resources is huge.Therefore,it is necessary to integrate the computer vision algorithm into the transmission line video monitoring system.This paper takes the construction crane near the transmission line as the research object,and takes crane arm touching wire during construction as the research behaviour,the key technologies used in the the monitoring process are analyzed,the main research contents are as follows:The algorithms of moving object detection are studied,and the Visual background extraction(Vibe)algorithm is selected for further research through comparative analysis in experiments.The shortcomings of Vibe algorithm are analyzed.The Vibe algorithm is improved by reconstructing background model,dynamic radius and connected domain analysis.The results of experiment show that compared with the original Vibe algorithm,the average F1 value of Improved Vibe algorithm is increased by 5.8%,and the average processing speed is decreased by 3.5%,the improvement of Vibe algorithm meets the expectation.The moving target tracking algorithms are studied,and Continuous adaptive meanshift(Camshift)algorithm is selected for further research through comparative analysis in experiments.For Camshift algorithm is easy to be affected by the color similarity between background and tracked target,an Improved Camshift algorithm based on the fusion of color features and texture features is proposed.The results of the experiment show that the Improved Camshift algorithm has a stable bhattacharyya distance around0.15,its effective frame rate is 84.5%,and compared with the original Camshift algorithm and Meanshift algorithm,its effective frame rate is increased by 47.3% and 11.3%,the tracking effect is significantly improved.This paper realizes the combination of transmission line anti external force damage monitoring interface and the above improved algorithms,chooses Py Qt5 frame and Qt Designer as the development tools of the monitoring interface,analyzes the requirements of the monitoring interface,and realizes each function module by programming.The construction behaviour of the crane is classified and the corresponding alarm level is set.The monitoring interface is tested and it has achieved all the functional requirements. |