[1]李名博,卫勇,穆志民,等.基于改进YOLOv7模型的朝天椒果实识别方法[J].江苏农业学报,2024,(12):2292-2301.[doi:doi:10.3969/j.issn.1000-4440.2024.12.012]
　LI Mingbo,WEI Yong,MU Zhimin,et al.Identification method of pod pepper fruits based on improved YOLOv7 model[J].,2024,(12):2292-2301.[doi:doi:10.3969/j.issn.1000-4440.2024.12.012]
点击复制

基于改进YOLOv7模型的朝天椒果实识别方法()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2024年12期
页码:
2292-2301
栏目:
农业信息工程
出版日期:
2024-12-30

文章信息/Info

Title:
Identification method of pod pepper fruits based on improved YOLOv7 model
作者:
李名博1; 卫勇1; 穆志民2; NASIR Mubarak Aliyu1
(1.天津农学院工程技术学院,天津300384；2.天津农学院基础科学学院,天津300384)
Author(s):
LI Mingbo1; WEI Yong1; MU Zhimin2; NASIR Mubarak Aliyu1
(1.College of Engineering and Technology, Tianjin Agricultural University, Tianjin 300384, China;2.College of Basic Sciences, Tianjin Agricultural University, Tianjin 300384, China)
关键词:
朝天椒; 果实识别; YOLOv7模型
Keywords:
pod pepper; fruit identification; YOLOv7 model
分类号:
S225;TP391;TP183
DOI:
doi:10.3969/j.issn.1000-4440.2024.12.012
文献标志码:
A
摘要:
朝天椒果实的准确识别是实现智能采摘的关键步骤,针对其生长环境复杂、果实大小不一、遮挡重叠等造成识别准确率低的问题,本研究提出一种基于改进YOLOv7的朝天椒果实识别方法。以YOLOv7为基础模型,设计了一种含有残差结构的AM_F模块,将其融合在YOLOv7主干网络中；基于空间、通道注意力机制的结构改进得到SAM_F、SE_ECA模块,分别将其融合在主干网络和颈部网络中,并进行结构简化，同时将SPP_CSP结构中的SPP替换为SPPF,实现参数计算量的精简,最终得到改进型YOLOv7模型——YOLOv7-F。采用对比试验对YOLOv7-F模型的识别效果进行验证分析,结果表明,YOLOv7-F模型对朝天椒果实的识别平均精度均值为80.07%,与YOLOv7模型相比,YOLOv7-F模型在识别时间加快23.4 ms的前提下,平均精度均值提升了1.06个百分点,而且模型大小也减少77.94 MB。YOLOv7-F模型实现了朝天椒果实识别精度和速度同步提升,为朝天椒果实智能采摘提供技术支撑。
Abstract:
The accurate identification of pod pepper fruits is the crucial step to realize intelligent picking. Aiming at the problem of low recognition accuracy caused by complex growing environments, different fruit sizes, and occlusion and overlapping, a fruit recognition method based on improved YOLOv7 was proposed. Using YOLOv7 as the basic model, an AM_F module with residual structure was designed and integrated into the backbone network of YOLOv7. The SAM_F and SE_ECA modules were obtained by improving the structure of the spatial and channel attention mechanisms. They were integrated into the backbone network and the neck network respectively, and the structure was simplified. At the same time, the SPP in the SPP_CSP structure was replaced by SPPF to simplify the calculation of parameters, and finally an improved YOLOv7 model, YOLOv7-F, was obtained. The recognition effect of YOLOv7-F model was verified and analyzed by comparison tests. The results indicated that the average recognition accuracy of YOLOv7-F model was 80.07%. Compared with the YOLOv7 model, the recognition time of the YOLOv7-F model was accelerated by 23.4 ms, the average accuracy was increased by 1.06 percentage points, and the model size was reduced by 77.94 MB. The YOLOv7-F model can realize the synchronous improvement of the recognition accuracy and recognition speed of pod pepper fruits, and provide technical support for the intelligent picking of pod pepper fruits.

参考文献/References:

[1]王立浩,张宝玺,张正海,等. “十三五”我国辣椒育种研究进展、产业现状及展望[J]. 中国蔬菜,2021(2):21-29.
[2]乔立娟,赵帮宏,宗义湘,等. 我国辣椒产业发展现状、趋势及对策[J]. 中国蔬菜,2023(11):9-15.
[3]王海楠,弋景刚,张秀花. 番茄采摘机器人识别与定位技术研究进展[J]. 中国农机化学报,2020,41(5):188-196.
[4]SANTOS T T, DE SOUZA L L, DOS SANTOS A A, et al. Grape detection, segmentation and tracking using deep neural networks and three-dimensional association[J]. Computers and Electronics in Agriculture,2020,170:105247.
[5]杨坚,钱振,张燕军,等.采用改进YOLOv4-tiny的复杂环境下番茄实时识别[J]. 农业工程学报,2022,38(9):215-221.
[6]HE Z X, KARKEE M, ZHANG Q. Detecting and localizing strawberry centers for robotic harvesting in field environment[J]. IFAC-PapersOnLine,2022,55(32):30-35.
[7]张楠楠,张晓,白铁成,等. 基于CBAM-YOLO v7的自然环境下棉叶病虫害识别方法[J]. 农业机械学报,2023,54(S1):239-244.
[8]王金星,马博,王震,等. 基于改进Mask R-CNN的苹果园害虫识别方法[J]. 农业机械学报,2023,54(6):253-263,360.
[9]JI W, GAO X X, XU B, et al. Target recognition method of green pepper harvesting robot based on manifold ranking[J]. Computers and Electronics in Agriculture,2020,177:105663.
[10]LI X, PAN J D, XIE F P, et al. Fast and accurate green pepper detection in complex backgrounds via an improved YOLOv4-tiny model[J]. Computers and Electronics in Agriculture,2021,191:106503.
[11]CONG P C, LI S D, ZHOU J C, et al. Research on instance segmentation algorithm of greenhouse sweet pepper detection based on improved mask RCNN[J]. Agronomy,2023,13(1):196.
[12]NAN Y L, ZHANG H C, ZENG Y, et al. Faster and accurate green pepper detection using NSGA-Ⅱ-based pruned YOLOv5l in the field environment[J]. Computers and Electronics in Agriculture,2023,205:107563.
[13]LI T H, SUN M, HE Q H, et al. Tomato recognition and location algorithm based on improved YOLOv5[J]. Computers and Electronics in Agriculture,2023,208:107759.
[14]MENG F, LI J, ZHANG Y, et al. Transforming unmanned pineapple picking with spatio-temporal convolutional neural networks[J]. Computers and Electronics in Agriculture,2023,214:108298.
[15]李恒,南新元,高丙朋,等. 一种基于GhostNet的绿色类圆果实识别方法[J]. 江苏农业学报,2023,39(3):724-731.
[16]ZHU W D, SUN J, WANG S M, et al. Segmentation and recognition of filed sweet pepper based on improved self-attention convolutional neural networks[J]. Multimedia Systems,2023,29(1):223-234.
[17]ZHONG S, XU W, ZHANG T, et al. Identification and Depth Localization of Clustered Pod Pepper Based on Improved Faster R-CNN[J]. IEEE Access,2022,10:93615-93625.
[18]LI D, SUN X, LV S, et al. A novel approach for the 3D localization of branch picking points based on deep learning applied to longan harvesting UAVs[J]. Computers and Electronics in Agriculture,2022,199:107191.
[19]翟先一,魏鸿磊,韩美奇,等. 基于改进YOLO卷积神经网络的水下海参检测[J]. 江苏农业学报,2023,39(7):1543-1553.
[20]王昱,姚兴智,李斌,等. 基于改进YOLO v7-tiny的甜椒畸形果识别算法[J]. 农业机械学报,2023,54(11):236-246.
[21]FU L, DUAN J, ZOU X, et al. Banana detection based on color and texture features in the natural environment[J]. Computers and Electronics in Agriculture,2019,167:105057.
[22]FANG W, WANG L, REN P. Tinier-YOLO: a real-time object detection method for constrained environments[J]. IEEE Access,2019,8:1935-1944.
[23]PARK K, HONG Y K, KIM G H, et al. Classification of apple leaf conditions in hyper-spectral images for diagnosis of Marssonina blotch using m RMR and deep neural network[J]. Computers and Electronics in Agriculture,2018,148:179-187.
[24]刘思幸,李爽,缪宏,等. 基于YOLOv3不同场景辣椒采摘机器人识别定位研究[J]. 农机化研究,2024,46(2):38-43.
[25]WU F, DUAN J, AI P, et al. Rachis detection and three-dimensional localization of cut off point for vision-based banana robot[J]. Computers and Electronics in Agriculture,2022,198:107079.
[26]JIN Y, YU C, YIN J, et al. Detection method for table grape ears and stems based on a far-close-range combined vision system and hand-eye-coordinated picking test[J]. Computers and Electronics in Agriculture,2022,202:107364.
[27]BAI Y H, MAO S H, ZHOU J, et al. Clustered tomato detection and picking point location using machine learning-aided image analysis for automatic robotic harvesting[J]. Precision Agriculture,2023,24(2):727-743.
[28]ZHAI S, SHANG D, WANG S, et al. DF-SSD: An improved SSD object detection algorithm based on Dense Net and feature fusion[J]. IEEE Access,2020,8:24344-24357.
[29]原昊. 基于深度学习的彩椒识别和定位技术研究[D]. 泰安：山东农业大学,2023.
[30]储鑫,李祥,罗斌,等. 基于改进YOLOv4算法的番茄叶部病害识别方法[J]. 江苏农业学报,2023,39(5):1199-1208.

相似文献/References:

[1]曾绍贵,朱邦彤,罗木旺,等.100份朝天椒的农艺性状和SRAP标记遗传多样性分析[J].江苏农业学报,2018,(04):871.[doi:doi:10.3969/j.issn.1000-4440.2018.04.023]
　ZENG Shao-gui,ZHU Bang-tong,LUO Mu-wang,et al.Agronomic characters and genetic diversity analysis of 100 pot pepper with SRAP markers[J].,2018,(12):871.[doi:doi:10.3969/j.issn.1000-4440.2018.04.023]
[2]魏茜雅,林欣琪,梁腊梅,等.褪黑素引发处理提高朝天椒种子萌发及幼苗耐盐性的生理机制[J].江苏农业学报,2022,38(06):1637.[doi:doi:10.3969/j.issn.1000-4440.2022.06.023]
　WEI Xi-ya,LIN Xin-qi,LIANG La-mei,et al.Physiological mechanism of melatonin soaking on improving seed germination and seedling salt tolerance of pepper[J].,2022,38(12):1637.[doi:doi:10.3969/j.issn.1000-4440.2022.06.023]
[3]方国文,何超,王鑫泽.基于YOLOv8n的轻量级巴旦木果实识别方法[J].江苏农业学报,2024,(09):1662.[doi:doi:10.3969/j.issn.1000-4440.2024.09.010]
　FANG Guowen,HE Chao,WANG Xinze.Lightweight almond fruit recognition method based on YOLOv8n[J].,2024,(12):1662.[doi:doi:10.3969/j.issn.1000-4440.2024.09.010]

备注/Memo

备注/Memo:
收稿日期：2024-04-05基金项目：黑龙江省重点研发计划项目(GA21C026)；天津市科技计划项目(21YDTPJC00600)作者简介：李名博(1998-),男,山西晋中人,硕士,主要从事智能农机装备研究。(E-mail)limingb257@163.com通讯作者：卫勇,(E-mail)weiytj@qq.com
更新日期/Last Update: 2025-01-23