A Roadmap for Recognizing Engineering Vehicle from Aerial Images of UAV

Abstract

Engineering vehicles on construction sites mainly include excavators, wheeled cranes and so on. If the engineering vehicle is working under or near the high-voltage power line, its bucket or boom probably enter the high-voltage breakdown range when they are lifted, which is very easy to result in accidents such as short circuit breakdown. So, it is necessary to find out the engineering vehicles working near high-voltage power line during inspection. Unmanned aerial vehicle (UAV) inspection is one of the main methods of electric power inspection at present. Lots of images are produced by UAV during the power line inspection. It will save a lot of inspection work if the engineering vehicles working near high-voltage power line can be recognized from these images. First, this paper analyzes the specific requirements of engineering vehicle recognition from aerial images of UAV power line inspection. Then, based on the research status of vehicle recognition in aerial images and other related fields at domestic and abroad, this paper comprehensively analyzes the research status of classical pattern recognition method and deep neural network method to recognize engineering vehicles in aerial images of UAV. Third, in view of the practical problems such as the low aerial image data of engineering vehicles, the roadmap of recognizing the engineering vehicles in the aerial image of UAV using the capsule network method is designed.