Autonomous UAV forced graffiti detection and removal system based on machine learning

This paper proposes a smart graffiti clean-up system based on an autonomous Unmanned Aerial Vehicle (UAV) platform. This smart clean system is based on edge detection and machine learning algorithms to realize the detection and tracking of graffiti image in real time. In Graffiti detection, we aim to build a model to detect graffiti on walls which can help navigate the UAV to the correct coordinate and estimate the area of the graffiti. The data set which contain graffiti images are trained using machine learning techniques which will be used for the detections of the graffiti patterns. This will automate the process of detecting the location of the graffiti based on the edge detection technique and the model will be able to estimate the area of the graffiti. To achieve obstacle detection, and collision, a smart navigation approach is also proposed with the help of LiDAR and external camera. The overall graffiti cleanup system contains hardware and software that allow the user to use spray enamel with the reach and scale of an autonomous UAV.