真实有雾场景下的目标检测  

Object Detection in Real-World Hazy Scene

作　　者：解宇虹 谢源 陈亮[1] 李翠华[1] 曲延云[1] Xie Yuhong;Xie Yuan;Chen Liang;Li Cuihua;Qu Yanyun(School of Information,Xiamen University,Xiamen 361000;School of Computer Science and Technology,East China Normal University,Shanghai 200062)

机构地区：[1]厦门大学信息学院,厦门361000 [2]华东师范大学计算机科学与技术学院,上海200062

出　　处：《计算机辅助设计与图形学学报》2021年第5期733-745,共13页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金面上项目(61876161,61772524);海峡联合基金(U1065252);北京市自然科学基金面上项目(4182067).

摘　　要：在有雾场景中实现对目标精确检测,是视频监控、智慧城市、无人驾驶等多个实际应用中一个重要的研究内容.为促进真实有雾场景下的目标检测研究,探讨了2个问题:有雾场景目标检测数据集的构建以及真实有雾场景下目标检测的解决方案.首先,设计了一种系统化的、具有真实感的有雾图像合成方法,并建立了合成有雾场景的目标检测数据集.同时,探讨了对真实有雾场景下目标检测器具有提升性能作用的数据集处理方法.其次,探讨了先验知识和模型的联合优化对真实有雾场景的目标检测性能的有效性,并提出了2个框架:基于知识引导的目标检测框架和基于图像去雾和目标检测的联合学习框架.基于知识引导的目标检测框架将统计先验知识用于指导通用目标检测网络学习有雾场景下的目标特征,使通用目标检测器能更好地适应特殊的目标检测场景.基于图像去雾和目标检测的联合学习框架通过去雾模型和目标检测模型的联合优化学习,有效学习图像去雾中恢复的结构细节和颜色特征,从而提高真实有雾场景下的目标检测精度.在RTTS数据集上的实验结果表明,基于知识引导的目标检测框架和基于图像去雾和目标检测的联合学习框架能够有效地提高有雾场景下目标检测器的性能,均值平均精度(mAP)分别为70.5%和66.6%.Accurate object detection in the real-world hazy scene is very important to some potential visual task,such as video surveillance,smart city,autonomous driving and so on.This paper focuses on two research problems,which are to build a synthetic dataset of object detection in hazy scene and to analyze the effect of prior knowledge and joint learning of model on object detection in real-world hazy scene.Two frameworks are proposed which are the knowledge-guided object detection(KODNet)and the joint learning in dehazing and object detection(DONet).In KODNet,statistical prior knowledge will be used to guide the general object detection network to learn object features in the hazy scene during the training,makes the general object detector better adapt to the special object detection scenario.DONet can effectively solve the problem of structural detail missing and color distortion caused by image dehazing,thereby realizing the improvement of the objects detection accuracy in real-world scene.The experimental results on RTTS show that KODNet and DONet are effective to the object detection in the real-world hazy scene and they achieve the mAP of 70.5%and 66.6% 。

关 键 词：真实有雾场景 目标检测 图像去雾 知识引导 联合学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]