Motion Target Detection and Recognition Based on YOLOv4 Algorithm

Abstract With the application of artificial intelligence more and widely, target detection “eyes” is becoming important, which can give machine ability to detect recognize image content in environment. In this paper, YOLOv4 deep learning motion algorithm used achieve positioning recognition targets, identify mark location type objects contained image, pictures, videos camera realtime detection, so that has most basic visual ability. Using algorithm, speed requirement real-time be realized with certain precision, model mAP value reached 69.48%.