摘要
目标检测是光学遥感图像分析的重要内容,是将图像数据转化为应用成果的关键一环。文章重点针对光学遥感图像中常见的飞机、舰船等目标,对目标检测技术进行了系统总结,并对未来重点发展方向进行了展望。重点论述的内容包括目标检测方法、目标检测性能评价准则和数据集,其中对于目标检测方法,从候选区域选择、特征学习、分类和后处理等四个方面进行了详细介绍。期望这些技术成果的总结分析和展望对目标检测技术的应用提供一定的借鉴和参考。
Object detection is an important part in optical remote sensing image analysis and a crucial linking between images and their applications. Taking the aircraft and ship research object, this paper systematically summarizes the latest researches and looks forward to the future development direction of object detection. The main content of this paper includes object detection methods, evaluation criteria and datasets. While introducing object detection methods in detail, this paper focusses on candidate region proposal, feature learning, classification and post-processing. It is expected that this survey can provide some references for the application of object detection in optical remote sensing images.
作者
李晓斌
江碧涛
杨渊博
傅雨泽
岳文振
LI Xiaobin;JIANG Bitao;YANG Yuanbo;FU Yuze;YUE Wenzhen(Beijing Institute of Remote Sensing Information, Beijing 100192, China;Department of Electronic Engineering, Tsinghua University, Beijing 100084, China)
出处
《航天返回与遥感》
CSCD
2019年第4期95-104,共10页
Spacecraft Recovery & Remote Sensing
关键词
目标检测方法
飞机检测
舰船检测
分类
遥感图像
object detection method
aircraft detection
ship detection
classification
remote sensing image
