UAV Detection System For Lifting Equipment

Lifting equipment is the most widely used construction site machinery.When a security accident of lifting equipment occurs,it may cause huge property losses and casualties.Therefore,the maintenance of lifting equipment is of great significance.The existing lifting equipment maintenance technology mainly relies on manpower,which suffers the problems of personnel danger,high requirement of maintenance technology,low efficiency and poor maintenance efficiency.With the advantages of safety and high maintenance efficiency,UAV has become one of the development trends of intelligent maintenance of lifting equipment.Driven by this point,this paper conducts major researches in four aspects: verticality defect detection of lifting equipment,defect location of lifting equipment,design of UAV detection hardware system,design of UAV detection software system:1)Verticality defect detection of lifting equipment.Aiming at solving the problem of high complexity and high security risk in traditional verticality defect detection methods,this section proposes an anomaly detection algorithm for verticality defect of lifting equipment based on UAV image.The algorithm uses the bottom and top images of the lifting equipment tower obtained by UAV and then converts these images to a same image spatial coordinate system.Through calculating the offsets of the two reference points,the proposed algorithm achieves the verticality result,which is then compared with the threshold of the safe verticality.This algorithm reduces the complexity of detection and improves the security of verticality defect detection.2)Defect location of lifting equipment.Aiming at the detect location of lifting equipment,this section proposes a new defect location algorithm of lifting equipment which combines the Siamese network and image registration technique.The proposed algorithm overcomes the problem that the features are not obvious.The algorithm firstly extracts the SIFT features of detail images and global images.Then the similarities among image blocks are compared by twin network.On this basis,the algorithm realizes the registration of detail image and global image.Therefore,the algorithm is able to get the defect location of lifting equipment for the follow-up maintenance.3)Design of UAV detection hardware system.Consider the specific scene of the lifting equipment detection,this section has designed overall structure of UAV detection hardware system,which includes accurate positioning module based on double-RTK positioning,obstacle avoidance module based on millimeter wave obstacle avoidance technology and camera module based on dual-head camera.4)Design of UAV detection software system.Based on the above methods,a system for crane defect analysis is designed.The application results show that the system owns the functions of system management,image display,basic drawing,equipment management,defect parameter acquisition and query statistics.Therefore,this system can assist the maintenance personnel to complete the lifting equipment defect detection work.