HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex Backgrounds
Apple leaf diseases are one of the most important factors that reduce apple quality and yield. The object detection technology based on deep learning can detect diseases in a timely manner and help automate disease control, thereby reducing economic losses. In the natural environment, tiny apple lea...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Plants |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2223-7747/12/15/2806 |
| _version_ | 1797586148084678656 |
|---|---|
| author | Xing Gao Zhiwen Tang Yubao Deng Shipeng Hu Hongmin Zhao Guoxiong Zhou |
| author_facet | Xing Gao Zhiwen Tang Yubao Deng Shipeng Hu Hongmin Zhao Guoxiong Zhou |
| author_sort | Xing Gao |
| collection | DOAJ |
| description | Apple leaf diseases are one of the most important factors that reduce apple quality and yield. The object detection technology based on deep learning can detect diseases in a timely manner and help automate disease control, thereby reducing economic losses. In the natural environment, tiny apple leaf disease targets (a resolution is less than 32 × 32 pixel<sup>2</sup>) are easily overlooked. To address the problems of complex background interference, difficult detection of tiny targets and biased detection of prediction boxes that exist in standard detectors, in this paper, we constructed a tiny target dataset TTALDD-4 containing four types of diseases, which include Alternaria leaf spot, Frogeye leaf spot, Grey spot and Rust, and proposed the HSSNet detector based on the YOLOv7-tiny benchmark for professional detection of apple leaf disease tiny targets. Firstly, the H-SimAM attention mechanism is proposed to focus on the foreground lesions in the complex background of the image. Secondly, SP-BiFormer Block is proposed to enhance the ability of the model to perceive tiny targets of leaf diseases. Finally, we use the SIOU loss to improve the case of prediction box bias. The experimental results show that HSSNet achieves 85.04% mAP (mean average precision), 67.53% AR (average recall), and 83 FPS (frames per second). Compared with other standard detectors, HSSNet maintains high real-time detection speed with higher detection accuracy. This provides a reference for the automated control of apple leaf diseases. |
| first_indexed | 2024-03-11T00:18:12Z |
| format | Article |
| id | doaj.art-2f4eef99797048149043cea5032072c7 |
| institution | Directory Open Access Journal |
| issn | 2223-7747 |
| language | English |
| last_indexed | 2024-03-11T00:18:12Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Plants |
| spelling | doaj.art-2f4eef99797048149043cea5032072c72023-11-18T23:26:07ZengMDPI AGPlants2223-77472023-07-011215280610.3390/plants12152806HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex BackgroundsXing Gao0Zhiwen Tang1Yubao Deng2Shipeng Hu3Hongmin Zhao4Guoxiong Zhou5College of Computer & Information Engineering, Central South University of Forestry and Technology, Changsha 410004, ChinaCollege of Computer & Information Engineering, Central South University of Forestry and Technology, Changsha 410004, ChinaCollege of Computer & Information Engineering, Central South University of Forestry and Technology, Changsha 410004, ChinaCollege of Computer & Information Engineering, Central South University of Forestry and Technology, Changsha 410004, ChinaCollege of Computer & Information Engineering, Central South University of Forestry and Technology, Changsha 410004, ChinaCollege of Computer & Information Engineering, Central South University of Forestry and Technology, Changsha 410004, ChinaApple leaf diseases are one of the most important factors that reduce apple quality and yield. The object detection technology based on deep learning can detect diseases in a timely manner and help automate disease control, thereby reducing economic losses. In the natural environment, tiny apple leaf disease targets (a resolution is less than 32 × 32 pixel<sup>2</sup>) are easily overlooked. To address the problems of complex background interference, difficult detection of tiny targets and biased detection of prediction boxes that exist in standard detectors, in this paper, we constructed a tiny target dataset TTALDD-4 containing four types of diseases, which include Alternaria leaf spot, Frogeye leaf spot, Grey spot and Rust, and proposed the HSSNet detector based on the YOLOv7-tiny benchmark for professional detection of apple leaf disease tiny targets. Firstly, the H-SimAM attention mechanism is proposed to focus on the foreground lesions in the complex background of the image. Secondly, SP-BiFormer Block is proposed to enhance the ability of the model to perceive tiny targets of leaf diseases. Finally, we use the SIOU loss to improve the case of prediction box bias. The experimental results show that HSSNet achieves 85.04% mAP (mean average precision), 67.53% AR (average recall), and 83 FPS (frames per second). Compared with other standard detectors, HSSNet maintains high real-time detection speed with higher detection accuracy. This provides a reference for the automated control of apple leaf diseases.https://www.mdpi.com/2223-7747/12/15/2806apple leaf diseasescomplex backgroundtiny-object detectionHSSNetTTALDD-4 |
| spellingShingle | Xing Gao Zhiwen Tang Yubao Deng Shipeng Hu Hongmin Zhao Guoxiong Zhou HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex Backgrounds Plants apple leaf diseases complex background tiny-object detection HSSNet TTALDD-4 |
| title | HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex Backgrounds |
| title_full | HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex Backgrounds |
| title_fullStr | HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex Backgrounds |
| title_full_unstemmed | HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex Backgrounds |
| title_short | HSSNet: A End-to-End Network for Detecting Tiny Targets of Apple Leaf Diseases in Complex Backgrounds |
| title_sort | hssnet a end to end network for detecting tiny targets of apple leaf diseases in complex backgrounds |
| topic | apple leaf diseases complex background tiny-object detection HSSNet TTALDD-4 |
| url | https://www.mdpi.com/2223-7747/12/15/2806 |
| work_keys_str_mv | AT xinggao hssnetaendtoendnetworkfordetectingtinytargetsofappleleafdiseasesincomplexbackgrounds AT zhiwentang hssnetaendtoendnetworkfordetectingtinytargetsofappleleafdiseasesincomplexbackgrounds AT yubaodeng hssnetaendtoendnetworkfordetectingtinytargetsofappleleafdiseasesincomplexbackgrounds AT shipenghu hssnetaendtoendnetworkfordetectingtinytargetsofappleleafdiseasesincomplexbackgrounds AT hongminzhao hssnetaendtoendnetworkfordetectingtinytargetsofappleleafdiseasesincomplexbackgrounds AT guoxiongzhou hssnetaendtoendnetworkfordetectingtinytargetsofappleleafdiseasesincomplexbackgrounds |