Research on the Application of Deep Learning Target Detection of Engineering Vehicles in the Patrol and Inspection for Military Optical Cable Lines by UAV

Abstract

The faulty-free operation of the optical cable lines is the key to the stable operation of the military optical cable communication network. Because the traditional manual hiking inspection and maintenance method for optical cable lines is not efficient, time-consuming and laborious, it is proposed to use the Unmanned Aerial Vehicle (UAV) to inspect the military optical cable lines. The deep learning Faster R-CNN target detection algorithm is applied to the identification of hidden dangers of the lines in the aerial image captured by the UAV. The hidden danger refers to the excavation damage of the engineering vehicle. Based on the Vehicle Detection in Aerial Imagery (VEDAI) dataset, the engineering vehicle aerial dataset was produced. Through simulation training, the target detection of excavators, bulldozers and other engineering vehicles was realized in the aerial image. The detection accuracy is better than traditional machine learning target detection algorithms such as Deformable Part Model (DPM) and Histogram of Oriented Gradient + Local Binary Pattern + Support Vector Machine (HOG+LBP+SVM). This research can provide some reference for the maintenance and inspection work of military optical cable lines by UAV.