Edge-Guided Camouflaged Object Detection via Multi-Level Feature Integration
Camouflaged object detection (COD) aims to segment those camouflaged objects that blend perfectly into their surroundings. Due to the low boundary contrast between camouflaged objects and their surroundings, their detection poses a significant challenge. Despite the numerous excellent camouflaged ob...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/13/5789 |
| _version_ | 1827734596419059712 |
|---|---|
| author | Kangwei Liu Tianchi Qiu Yinfeng Yu Songlin Li Xiuhong Li |
| author_facet | Kangwei Liu Tianchi Qiu Yinfeng Yu Songlin Li Xiuhong Li |
| author_sort | Kangwei Liu |
| collection | DOAJ |
| description | Camouflaged object detection (COD) aims to segment those camouflaged objects that blend perfectly into their surroundings. Due to the low boundary contrast between camouflaged objects and their surroundings, their detection poses a significant challenge. Despite the numerous excellent camouflaged object detection methods developed in recent years, issues such as boundary refinement and multi-level feature extraction and fusion still need further exploration. In this paper, we propose a novel multi-level feature integration network (MFNet) for camouflaged object detection. Firstly, we design an edge guidance module (EGM) to improve the COD performance by providing additional boundary semantic information by combining high-level semantic information and low-level spatial details to model the edges of camouflaged objects. Additionally, we propose a multi-level feature integration module (MFIM), which leverages the fine local information of low-level features and the rich global information of high-level features in adjacent three-level features to provide a supplementary feature representation for the current-level features, effectively integrating the full context semantic information. Finally, we propose a context aggregation refinement module (CARM) to efficiently aggregate and refine the cross-level features to obtain clear prediction maps. Our extensive experiments on three benchmark datasets show that the MFNet model is an effective COD model and outperforms other state-of-the-art models in all four evaluation metrics (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>S</mi><mi>α</mi></msub></semantics></math></inline-formula>, <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>E</mi><mi>ϕ</mi></msub></semantics></math></inline-formula>, <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msubsup><mi>F</mi><mi>β</mi><mi>w</mi></msubsup></semantics></math></inline-formula>, and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>M</mi><mi>A</mi><mi>E</mi></mrow></semantics></math></inline-formula>). |
| first_indexed | 2024-03-11T01:29:49Z |
| format | Article |
| id | doaj.art-e9f0455e59504c7a8c1e9b5bc07d0407 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T01:29:49Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-e9f0455e59504c7a8c1e9b5bc07d04072023-11-18T17:26:46ZengMDPI AGSensors1424-82202023-06-012313578910.3390/s23135789Edge-Guided Camouflaged Object Detection via Multi-Level Feature IntegrationKangwei Liu0Tianchi Qiu1Yinfeng Yu2Songlin Li3Xiuhong Li4Key Laboratory of Signal Detection and Processing, Department of Information Science and Engineering, Xinjiang University, Urumqi 830017, ChinaKey Laboratory of Signal Detection and Processing, Department of Information Science and Engineering, Xinjiang University, Urumqi 830017, ChinaKey Laboratory of Signal Detection and Processing, Department of Information Science and Engineering, Xinjiang University, Urumqi 830017, ChinaKey Laboratory of Signal Detection and Processing, Department of Information Science and Engineering, Xinjiang University, Urumqi 830017, ChinaKey Laboratory of Signal Detection and Processing, Department of Information Science and Engineering, Xinjiang University, Urumqi 830017, ChinaCamouflaged object detection (COD) aims to segment those camouflaged objects that blend perfectly into their surroundings. Due to the low boundary contrast between camouflaged objects and their surroundings, their detection poses a significant challenge. Despite the numerous excellent camouflaged object detection methods developed in recent years, issues such as boundary refinement and multi-level feature extraction and fusion still need further exploration. In this paper, we propose a novel multi-level feature integration network (MFNet) for camouflaged object detection. Firstly, we design an edge guidance module (EGM) to improve the COD performance by providing additional boundary semantic information by combining high-level semantic information and low-level spatial details to model the edges of camouflaged objects. Additionally, we propose a multi-level feature integration module (MFIM), which leverages the fine local information of low-level features and the rich global information of high-level features in adjacent three-level features to provide a supplementary feature representation for the current-level features, effectively integrating the full context semantic information. Finally, we propose a context aggregation refinement module (CARM) to efficiently aggregate and refine the cross-level features to obtain clear prediction maps. Our extensive experiments on three benchmark datasets show that the MFNet model is an effective COD model and outperforms other state-of-the-art models in all four evaluation metrics (<inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>S</mi><mi>α</mi></msub></semantics></math></inline-formula>, <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>E</mi><mi>ϕ</mi></msub></semantics></math></inline-formula>, <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msubsup><mi>F</mi><mi>β</mi><mi>w</mi></msubsup></semantics></math></inline-formula>, and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>M</mi><mi>A</mi><mi>E</mi></mrow></semantics></math></inline-formula>).https://www.mdpi.com/1424-8220/23/13/5789camouflaged object detectionmulti-level feature integrationattention mechanismboundary semantic information |
| spellingShingle | Kangwei Liu Tianchi Qiu Yinfeng Yu Songlin Li Xiuhong Li Edge-Guided Camouflaged Object Detection via Multi-Level Feature Integration Sensors camouflaged object detection multi-level feature integration attention mechanism boundary semantic information |
| title | Edge-Guided Camouflaged Object Detection via Multi-Level Feature Integration |
| title_full | Edge-Guided Camouflaged Object Detection via Multi-Level Feature Integration |
| title_fullStr | Edge-Guided Camouflaged Object Detection via Multi-Level Feature Integration |
| title_full_unstemmed | Edge-Guided Camouflaged Object Detection via Multi-Level Feature Integration |
| title_short | Edge-Guided Camouflaged Object Detection via Multi-Level Feature Integration |
| title_sort | edge guided camouflaged object detection via multi level feature integration |
| topic | camouflaged object detection multi-level feature integration attention mechanism boundary semantic information |
| url | https://www.mdpi.com/1424-8220/23/13/5789 |
| work_keys_str_mv | AT kangweiliu edgeguidedcamouflagedobjectdetectionviamultilevelfeatureintegration AT tianchiqiu edgeguidedcamouflagedobjectdetectionviamultilevelfeatureintegration AT yinfengyu edgeguidedcamouflagedobjectdetectionviamultilevelfeatureintegration AT songlinli edgeguidedcamouflagedobjectdetectionviamultilevelfeatureintegration AT xiuhongli edgeguidedcamouflagedobjectdetectionviamultilevelfeatureintegration |