摘要
目标检测是计算机视觉领域中的核心课题之一。其主要任务是从图像信息中检测并定位特定的目标。卷积神经网络是目标检测的常用方法,文章重点阐述了YOLO算法的实现原理和结构设计,从传统的目标检测过渡到二阶段检测算法。
Object detection is one of the core subjects in the field of computer vision.Its main task is to detect and locate specific targets from the image information.Convolutional neural network is a common method of target detection.This paper focuses on the realization principle and structure design of YOLO algorithm,and the transition from traditional target detection to two-stage detection algorithm.
作者
徐畅
程文轩
杨远舟
XU Chang;CHENG Wen-xuan;YANG Yuan-zhou(Chongqing University of Posts and Telecommunications,Chongqing 400065,China;University of Jinan,Jinan,Shandong 250022,China)
出处
《电脑与信息技术》
2020年第4期45-47,共3页
Computer and Information Technology