Pattern Recognition Through Digital Image Processing for Unmanned Aerial Vehicles

This paper describes the implementation of two digital image processing methods for pattern recognition, by color boundary method and the Haar Cascade Classifier to detect objects in a video stream, both methods implemented on Python 3 and OpenCV. Patterns detection of images obtained from drones has advantages over traditional video recording drones. The drone has a streaming video system, based on the Raspberry Pi 3 minicomputer, which is sent by wireless communication to the base station where a pattern recognition algorithm performs operations on the video source coming from the drone. This proposed system has excellent performance based on an integrated streaming video system with 5.8GHz Wi-Fi connection acceptable to the base station. Both methods proved to be valid for certain types of patterns and objects under different light conditions.