Based on the improved YOLOV3 small target detection algorithm

Abstract

Target detection is one of the core problems in the field of machine vision. However, the different appearance, shape, and size of objects and the influence of interference factors such as illumination and occlusion during the imaging process pose a considerable challenge to the task of target detection. This article introduces YOLOV3 network structure and related research, proposes a small target detection algorithm based on improved YOLOV3, describes the flow of YOLOV3 model to achieve small target detection algorithm, analyzes the experiment and results, and summarizes the problems in the implementation process and shortcomings, provide a reference for the continuous improvement of small target detection in the future.