一种基于图像目标检测的无人机相机精准对焦拍照方法  

Accurate Focus Photography Method for Drone Camera Based on Image Target Detection

作　　者：刘高 饶成成 王昊 LIU Gao;RAO Cheng-cheng;WANG Hao(Guangdong Power Grid Co.,Ltd.,Machine Patrol Management Center,Guangzhou Guangdong 510163;Beijing Chenical Industry Group Co.,Ltd.,Guangzhou Guangdong 510700)

机构地区：[1]广东电网有限责任公司机巡管理中心,广东广州510163 [2]南方电网数字电网研究院有限公司,广东广州510700

出　　处：《数字技术与应用》2020年第12期99-101,共3页Digital Technology & Application

基　　金：基于AI图像识别的油电混动无人机高精度自动驾驶技术研究:GDKJXM20184730(036100KK52180003)。

摘　　要：为提高对输电线路巡检的密度和精度,降低输电线路的运维成本,无人机巡检技术被引入到了电力系统巡检业务中。本文基于图像目标检测的无人机相机精准对焦拍照方法,依据实验论证该方法能辅助无人机飞行避障,提升了拍摄的图片质量。In order to improve the density and accuracy of transmission line inspections and reduce the operation and maintenance costs of transmission lines,drone inspection technology has been introduced into the power system inspection business.In this paper,the precise focus photography method of UAV camera based on image target detection,based on experimental demonstration,this method can assist UAV flight avoidance and improve the picture quality.

关 键 词：无人机巡检 深度学习 SSD模型 云台控制 自动对焦 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]