Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5
Rapid and precise detection of cucumbers is a key element in enhancing the capability of intelligent harvesting robots. Problems such as near-color background interference, branch and leaf occlusion of fruits, and target scale diversity in greenhouse environments posed higher requirements for cucumb...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-08-01
|
Series: | Agronomy |
Subjects: | |
Online Access: | https://www.mdpi.com/2073-4395/13/8/2062 |
_version_ | 1797585854035656704 |
---|---|
author | Liyang Su Haixia Sun Shujuan Zhang Xinyuan Lu Runrun Wang Linjie Wang Ning Wang |
author_facet | Liyang Su Haixia Sun Shujuan Zhang Xinyuan Lu Runrun Wang Linjie Wang Ning Wang |
author_sort | Liyang Su |
collection | DOAJ |
description | Rapid and precise detection of cucumbers is a key element in enhancing the capability of intelligent harvesting robots. Problems such as near-color background interference, branch and leaf occlusion of fruits, and target scale diversity in greenhouse environments posed higher requirements for cucumber target detection algorithms. Therefore, a lightweight YOLOv5s-Super model was proposed based on the YOLOv5s model. First, in this study, the bidirectional feature pyramid network (BiFPN) and C3CA module were added to the YOLOv5s-Super model with the goal of capturing cucumber shoulder features of long-distance dependence and dynamically fusing multi-scale features in the near-color background. Second, the Ghost module was added to the YOLOv5s-Super model to speed up the inference time and floating-point computation speed of the model. Finally, this study visualized different feature fusion methods for the BiFPN module; independently designed a C3SimAM module for comparison between parametric and non-parametric attention mechanisms. The results showed that the YOLOv5s-Super model achieves mAP of 87.5%, which was 4.2% higher than the YOLOv7-tiny and 1.9% higher than the YOLOv8s model. The improved model could more accurately and robustly complete the detection of multi-scale features in complex near-color backgrounds while the model met the requirement of being lightweight. These results could provide technical support for the implementation of intelligent cucumber picking. |
first_indexed | 2024-03-11T00:13:00Z |
format | Article |
id | doaj.art-f122f425032f4eaca1d00fda74ca75bb |
institution | Directory Open Access Journal |
issn | 2073-4395 |
language | English |
last_indexed | 2024-03-11T00:13:00Z |
publishDate | 2023-08-01 |
publisher | MDPI AG |
record_format | Article |
series | Agronomy |
spelling | doaj.art-f122f425032f4eaca1d00fda74ca75bb2023-11-18T23:54:30ZengMDPI AGAgronomy2073-43952023-08-01138206210.3390/agronomy13082062Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5Liyang Su0Haixia Sun1Shujuan Zhang2Xinyuan Lu3Runrun Wang4Linjie Wang5Ning Wang6College of Agricultural Engineering, Shanxi Agricultural University, Jinzhong 030801, ChinaCollege of Agricultural Engineering, Shanxi Agricultural University, Jinzhong 030801, ChinaCollege of Agricultural Engineering, Shanxi Agricultural University, Jinzhong 030801, ChinaCollege of Agricultural Engineering, Shanxi Agricultural University, Jinzhong 030801, ChinaCollege of Agricultural Engineering, Shanxi Agricultural University, Jinzhong 030801, ChinaCollege of Agricultural Engineering, Shanxi Agricultural University, Jinzhong 030801, ChinaDepartment of Biosystems and Agricultural Engineering, Oklahoma State University, 111 Agricultural Hall, Stillwater, OK 74078, USARapid and precise detection of cucumbers is a key element in enhancing the capability of intelligent harvesting robots. Problems such as near-color background interference, branch and leaf occlusion of fruits, and target scale diversity in greenhouse environments posed higher requirements for cucumber target detection algorithms. Therefore, a lightweight YOLOv5s-Super model was proposed based on the YOLOv5s model. First, in this study, the bidirectional feature pyramid network (BiFPN) and C3CA module were added to the YOLOv5s-Super model with the goal of capturing cucumber shoulder features of long-distance dependence and dynamically fusing multi-scale features in the near-color background. Second, the Ghost module was added to the YOLOv5s-Super model to speed up the inference time and floating-point computation speed of the model. Finally, this study visualized different feature fusion methods for the BiFPN module; independently designed a C3SimAM module for comparison between parametric and non-parametric attention mechanisms. The results showed that the YOLOv5s-Super model achieves mAP of 87.5%, which was 4.2% higher than the YOLOv7-tiny and 1.9% higher than the YOLOv8s model. The improved model could more accurately and robustly complete the detection of multi-scale features in complex near-color backgrounds while the model met the requirement of being lightweight. These results could provide technical support for the implementation of intelligent cucumber picking.https://www.mdpi.com/2073-4395/13/8/2062near-color backgroundtarget detectionlightweight algorithmattention mechanismfeature fusion |
spellingShingle | Liyang Su Haixia Sun Shujuan Zhang Xinyuan Lu Runrun Wang Linjie Wang Ning Wang Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5 Agronomy near-color background target detection lightweight algorithm attention mechanism feature fusion |
title | Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5 |
title_full | Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5 |
title_fullStr | Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5 |
title_full_unstemmed | Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5 |
title_short | Cucumber Picking Recognition in Near-Color Background Based on Improved YOLOv5 |
title_sort | cucumber picking recognition in near color background based on improved yolov5 |
topic | near-color background target detection lightweight algorithm attention mechanism feature fusion |
url | https://www.mdpi.com/2073-4395/13/8/2062 |
work_keys_str_mv | AT liyangsu cucumberpickingrecognitioninnearcolorbackgroundbasedonimprovedyolov5 AT haixiasun cucumberpickingrecognitioninnearcolorbackgroundbasedonimprovedyolov5 AT shujuanzhang cucumberpickingrecognitioninnearcolorbackgroundbasedonimprovedyolov5 AT xinyuanlu cucumberpickingrecognitioninnearcolorbackgroundbasedonimprovedyolov5 AT runrunwang cucumberpickingrecognitioninnearcolorbackgroundbasedonimprovedyolov5 AT linjiewang cucumberpickingrecognitioninnearcolorbackgroundbasedonimprovedyolov5 AT ningwang cucumberpickingrecognitioninnearcolorbackgroundbasedonimprovedyolov5 |