A new benchmark for camouflaged object detection: RGB-D camouflaged object detection dataset

IF 1.8 4区物理与天体物理 Q2 PHYSICS, MULTIDISCIPLINARY

Open Physics Pub Date : 2024-07-20 DOI:10.1515/phys-2024-0060

Dongdong Zhang, Chunping Wang, Qiang Fu

{"title":"A new benchmark for camouflaged object detection: RGB-D camouflaged object detection dataset","authors":"Dongdong Zhang, Chunping Wang, Qiang Fu","doi":"10.1515/phys-2024-0060","DOIUrl":null,"url":null,"abstract":"This article aims to provide a novel image paradigm for camouflaged object detection, i.e., RGB-D images. To promote the development of camouflaged object detection tasks based on RGB-D images, we construct an RGB-D camouflaged object detection dataset, dubbed CODD. This dataset is obtained by converting the existing salient object detection RGB-D datasets by image-to-image translation techniques, which is comparable to the current widely used camouflaged object detection dataset in terms of diversity and complexity. In particular, in order to obtain high-quality translated images, we design a selection strategy that takes into account the structural similarity between pre- and post-conversion images, the similarity between the appearance of objects and their surroundings, as well as the ambiguity of object boundaries. In addition, we extensively evaluate the CODD dataset using existing RGB-D-based salient object detection methods to validate the challenge and usability of the dataset. The CODD dataset will be available at: https://github.com/zcc0616/CODD-Dateset.git.","PeriodicalId":48710,"journal":{"name":"Open Physics","volume":null,"pages":null},"PeriodicalIF":1.8000,"publicationDate":"2024-07-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Open Physics","FirstCategoryId":"101","ListUrlMain":"https://doi.org/10.1515/phys-2024-0060","RegionNum":4,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"PHYSICS, MULTIDISCIPLINARY","Score":null,"Total":0}

引用次数: 0

Abstract

This article aims to provide a novel image paradigm for camouflaged object detection, i.e., RGB-D images. To promote the development of camouflaged object detection tasks based on RGB-D images, we construct an RGB-D camouflaged object detection dataset, dubbed CODD. This dataset is obtained by converting the existing salient object detection RGB-D datasets by image-to-image translation techniques, which is comparable to the current widely used camouflaged object detection dataset in terms of diversity and complexity. In particular, in order to obtain high-quality translated images, we design a selection strategy that takes into account the structural similarity between pre- and post-conversion images, the similarity between the appearance of objects and their surroundings, as well as the ambiguity of object boundaries. In addition, we extensively evaluate the CODD dataset using existing RGB-D-based salient object detection methods to validate the challenge and usability of the dataset. The CODD dataset will be available at: https://github.com/zcc0616/CODD-Dateset.git.

查看原文本刊更多论文

伪装物体检测的新基准：RGB-D 伪装物体检测数据集

本文旨在为伪装物体检测提供一种新的图像范例，即 RGB-D 图像。为了促进基于 RGB-D 图像的伪装物体检测任务的发展，我们构建了一个 RGB-D 伪装物体检测数据集，命名为 CODD。该数据集是通过图像到图像转换技术将现有的突出物体检测 RGB-D 数据集转换而来，在多样性和复杂性方面与目前广泛使用的伪装物体检测数据集相当。特别是，为了获得高质量的翻译图像，我们设计了一种选择策略，该策略考虑了转换前和转换后图像的结构相似性、物体外观与其周围环境的相似性以及物体边界的模糊性。此外，我们还利用现有的基于 RGB-D 的突出物体检测方法对 CODD 数据集进行了广泛评估，以验证该数据集的挑战性和可用性。CODD 数据集可在以下网址获取：https://github.com/zcc0616/CODD-Dateset.git。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Open Physics PHYSICS, MULTIDISCIPLINARY-

CiteScore

3.20

自引率

5.30%

发文量

审稿时长

18 weeks

期刊介绍： Open Physics is a peer-reviewed, open access, electronic journal devoted to the publication of fundamental research results in all fields of physics. The journal provides the readers with free, instant, and permanent access to all content worldwide; and the authors with extensive promotion of published articles, long-time preservation, language-correction services, no space constraints and immediate publication. Our standard policy requires each paper to be reviewed by at least two Referees and the peer-review process is single-blind.