Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter
The Segment Anything Model (SAM) is a versatile image segmentation model that enables zero-shot segmentation of various objects in any image using prompts, including bounding boxes, points, texts, and more. However, studies have shown that the SAM performs poorly in agricultural tasks like crop dise...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-09-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/18/7884 |
| _version_ | 1797576979329843200 |
|---|---|
| author | Yaqin Li Dandan Wang Cao Yuan Hao Li Jing Hu |
| author_facet | Yaqin Li Dandan Wang Cao Yuan Hao Li Jing Hu |
| author_sort | Yaqin Li |
| collection | DOAJ |
| description | The Segment Anything Model (SAM) is a versatile image segmentation model that enables zero-shot segmentation of various objects in any image using prompts, including bounding boxes, points, texts, and more. However, studies have shown that the SAM performs poorly in agricultural tasks like crop disease segmentation and pest segmentation. To address this issue, the agricultural SAM adapter (ASA) is proposed, which incorporates agricultural domain expertise into the segmentation model through a simple but effective adapter technique. By leveraging the distinctive characteristics of agricultural image segmentation and suitable user prompts, the model enables zero-shot segmentation, providing a new approach for zero-sample image segmentation in the agricultural domain. Comprehensive experiments are conducted to assess the efficacy of the ASA compared to the default SAM. The results show that the proposed model achieves significant improvements on all 12 agricultural segmentation tasks. Notably, the average Dice score improved by 41.48% on two coffee-leaf-disease segmentation tasks. |
| first_indexed | 2024-03-10T22:02:23Z |
| format | Article |
| id | doaj.art-6f5b414ec50746939d15af2f29a3a52b |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T22:02:23Z |
| publishDate | 2023-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-6f5b414ec50746939d15af2f29a3a52b2023-11-19T12:55:33ZengMDPI AGSensors1424-82202023-09-012318788410.3390/s23187884Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model AdapterYaqin Li0Dandan Wang1Cao Yuan2Hao Li3Jing Hu4School of Mathematics and Computer Science, Wuhan Polytechnic University, Wuhan 430024, ChinaSchool of Mathematics and Computer Science, Wuhan Polytechnic University, Wuhan 430024, ChinaSchool of Mathematics and Computer Science, Wuhan Polytechnic University, Wuhan 430024, ChinaSchool of Mathematics and Computer Science, Wuhan Polytechnic University, Wuhan 430024, ChinaSchool of Mathematics and Computer Science, Wuhan Polytechnic University, Wuhan 430024, ChinaThe Segment Anything Model (SAM) is a versatile image segmentation model that enables zero-shot segmentation of various objects in any image using prompts, including bounding boxes, points, texts, and more. However, studies have shown that the SAM performs poorly in agricultural tasks like crop disease segmentation and pest segmentation. To address this issue, the agricultural SAM adapter (ASA) is proposed, which incorporates agricultural domain expertise into the segmentation model through a simple but effective adapter technique. By leveraging the distinctive characteristics of agricultural image segmentation and suitable user prompts, the model enables zero-shot segmentation, providing a new approach for zero-sample image segmentation in the agricultural domain. Comprehensive experiments are conducted to assess the efficacy of the ASA compared to the default SAM. The results show that the proposed model achieves significant improvements on all 12 agricultural segmentation tasks. Notably, the average Dice score improved by 41.48% on two coffee-leaf-disease segmentation tasks.https://www.mdpi.com/1424-8220/23/18/7884image segmentationadaptersagricultural image segmentationzero-shot segmentation |
| spellingShingle | Yaqin Li Dandan Wang Cao Yuan Hao Li Jing Hu Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter Sensors image segmentation adapters agricultural image segmentation zero-shot segmentation |
| title | Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter |
| title_full | Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter |
| title_fullStr | Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter |
| title_full_unstemmed | Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter |
| title_short | Enhancing Agricultural Image Segmentation with an Agricultural Segment Anything Model Adapter |
| title_sort | enhancing agricultural image segmentation with an agricultural segment anything model adapter |
| topic | image segmentation adapters agricultural image segmentation zero-shot segmentation |
| url | https://www.mdpi.com/1424-8220/23/18/7884 |
| work_keys_str_mv | AT yaqinli enhancingagriculturalimagesegmentationwithanagriculturalsegmentanythingmodeladapter AT dandanwang enhancingagriculturalimagesegmentationwithanagriculturalsegmentanythingmodeladapter AT caoyuan enhancingagriculturalimagesegmentationwithanagriculturalsegmentanythingmodeladapter AT haoli enhancingagriculturalimagesegmentationwithanagriculturalsegmentanythingmodeladapter AT jinghu enhancingagriculturalimagesegmentationwithanagriculturalsegmentanythingmodeladapter |