摘要
针对老照片档案存在局部画面模糊不清,甚至残缺而导致目标特征不明显等问题,运用基于YOLOv4的图像检测技术对其进行有效检测实验,发现能够实现精准识别和检出。该技术可为探索老照片档案智能化有效检索提供启发和借鉴。
Aiming at the problems of blurred or even incomplete images in old photo archives,which lead to inconspicuous target features,this paper uses Image detection technology based on YOLOv4 to conduct effective detection experiments on them,and finds that accurate identification and detection can be achieved.This technology can provide inspiration and reference for the exploration of intelligent and effective retrieval of old photo archives.
作者
徐辛酉
杨晓芳
陈文康
XU Xin-you;YANG Xiao-fang;CHEN Wen-kang(College of Management,Guangxi Minzu University,Guangxi 530006;College of Computer Science&Information Engineering,Guangxi Normal University,Guangxi 541004)
出处
《山西档案》
2022年第5期150-156,共7页
Shanxi Archives
基金
国家社科基金项目“新《档案法》规制下档案数字监管模式研究”(项目编号:22BTQ091)
广西千骨计划人文社科类立项课题“面向社会风险防控的建档理论构建与实现研究”(编号:2021QGRW034)阶段性研究成果。
