UAV image and intelligent detection of building surface cracks

To solve the problems of low efficiency and high risk of crack detection in high-rise buildings, the author proposes a building crack image acquisition and processing system based on unmanned aerial vehicle (UAV), expounds the working principle of the system, and analyzes the selection of hardware in the system. The experiment is carried out by combining the acquisition system with the image processing system of the ground station. Based on the characteristics of building surface crack image, an adaptive filtering method is introduced to improve the classic system file checker combination method. The experimental results show that the accuracy of Canny’s method was 96.88%. Canny-based crack target detection method is trained and tested on the crack data set to obtain the classification results and the best detection effect. It is proved that the Canny algorithm based on UAV images can effectively detect cracks on building surface.