基于改进YOLOV3的自然环境下绿色柑橘的识别算法*

doi:10.13733/j.jcam.issn.20955553.2021.11.24

中国农机化学报 ›› 2021, Vol. 42 ›› Issue (11): 159-165.DOI: 10.13733/j.jcam.issn.20955553.2021.11.24

基于改进YOLOV3的自然环境下绿色柑橘的识别算法^*

宋中山^1,2, 刘越^1,2, 郑禄^1,2, 帖军^1,2, 汪进^1,2

1.中南民族大学计算机科学学院,武汉市,430074;
2.湖北省制造企业智能管理工程技术研究中心,武汉市,430074

收稿日期:2020-12-08 修回日期:2021-01-26 出版日期:2021-11-15 发布日期:2021-11-15
通讯作者: 郑禄,男,1989年生,内蒙古乌兰察布人,实验师;研究方向为深度学习与图形识别。E-mail: lu2008@mail.scuec.edu.cn
作者简介:宋中山,男,1963年生,湖北仙桃人,副教授;研究方向为图像处理与模式识别。E-mail: songzs@mail.scuec.edu.cn
基金资助:
*湖北省技术创新专项重大项目(2019ABA101);中国科学院—国家民委农业信息技术研究与开发联合实验室招标课题(PJW060012003);中央高校基本科研业务费专项资金项目(CZT19012)

Identification of green citrus based on improved YOLOV3 in natural environment

Song Zhongshan^1,2, Liu Yue^1,2, Zheng Lu^1,2, Tie Jun^1,2, Wang Jin^1,2

1. College of Computer Science, South-Central University for Nationalities, Wuhan, 430074, China;
2. Hubei Provincial Engineering Research Center for Intelligent Management of Manufacturing Enterprises, Wuhan, 430074, China

Received:2020-12-08 Revised:2021-01-26 Online:2021-11-15 Published:2021-11-15

摘要/Abstract

摘要： 为研究自然环境下柑橘的图像识别技术,实现柑橘的早期产量预测,提出一种改进的D-YOLOV3算法,实现自然环境下未成熟的绿色柑橘的识别与检测。研究构建绿色柑橘图像数据集,并对采集的图像进行预处理;改进算法采用DenseNet的密集连接机制替换YOLOV3网络中的特征提取网络Darknet53中的后三个下采样层,加强特征的传播,实现特征的复用。通过自制的数据集对D-YOLOV3算法进行测试,并分别对改进前后网络的识别性能、不同预处理方法和不同数据量图像对模型的影响进行试验。试验结果表明,改进的D-YOLOV3算法相对于传统YOLOV3算法精确率提高6.57%,召回率提高2.75%,F₁分数提高4.41%,交并比提高6.13%,平均单张检测时间为0.28 s。通过不同果实数量图像对比试验验证了算法的可行性和准确性。研究结果表明,本文提出的D-YOLOV3算法对自然环境下未成熟的绿色柑橘识别具有较高的精准度,为柑橘的早期测产提供了技术支持。

关键词: 目标检测, YOLOV3算法, DenseNet算法, 绿色柑橘

Abstract: In order to study the image recognition technology of citrus in natural environment and realize the early yield prediction of citrus, an improved D-YOLOV3 algorithm was proposed. In this study, a green citrus image data set was constructed. In order to enhance the diversity of the data set, preprocessing operations were carried out on the collected images, including color balance, brightness transformation, rotation transformation, blur, and noise. To solve the problem that gradient information in deep networks will disappear or over-expand with the deepening of the network, the improved model adopts DenseNet's dense connection mechanism to replace the last three lower sampling layers of feature extraction network Darknet53 in the YOLOV3 network to enhance the propagation of features and realize feature reuse. The D-YOLOV3 model was tested by the self-made data set, and experiments were conducted on the recognition performance of the network before and after the modification, different pretreatment methods, different amounts of fruit, and different amounts of data images on the model. The experimental results show that compared with the traditional YOLOV3 model, the accuracy rate of the improved D-YOLOV3 model is increased by 6.57%, the recall rate is increased by 2.75%, the F₁ score is increased by 4.41%, the intersection ratio is increased by 6.13%, and the average single test time is 0.28 s. Different preprocessing methods enhance the robustness of the model, among which the fuzzy processing has the greatest influence on the performance of the model and the rotation change has the least influence. In the multi-fruit scene, the improved model has a higher recognition accuracy of 5.53% than before, which proves that the model has advantages in recognizing multi-target fruit in the actual scene. Only 1,250 images are needed to fit the model. The research results show that the D-YOLOV3 model proposed in this paper has high accuracy in recognizing immature green citrus in the natural environment, providing technical support for the early production measurement of citrus.

Key words: object detection, YOLOV3 algorithm, DenseNet algorithm, green citrus

中图分类号:

TP391.4

宋中山, 刘越, 郑禄, 帖军, 汪进. 基于改进YOLOV3的自然环境下绿色柑橘的识别算法^*[J]. 中国农机化学报, 2021, 42(11): 159-165.

Song Zhongshan, Liu Yue, Zheng Lu, Tie Jun, Wang Jin. Identification of green citrus based on improved YOLOV3 in natural environment[J]. Journal of Chinese Agricultural Mechanization, 2021, 42(11): 159-165.

参考文献

[1] 沈兆敏. 世界柑橘产销现状及做强我国柑橘产业的建议[J]. 果农之友, 2020(3): 1-3.
[2] 程洪, Lutz D, Michael B, 等. 基于图像处理与支持向量机的树上苹果早期估产研究[J]. 农业机械学报, 2015, 46(3): 9-14, 22.
[3] 黄小玉, 李光林, 马驰, 等. 基于改进判别区域特征融合算法的近色背景绿色桃子识别[J].农业工程学报, 2018, 34(23): 142-148.
Huang Xiaoyu, Li Guanglin, Ma Chi, et al. Green peach recognition based on improved discriminative regional feature integration algorithm in similar background [J]. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(23): 142-148.
[4] 闫建伟, 赵源, 张乐伟, 等. 改进Faster-RCNN自然环境下识别刺梨果实[J]. 农业工程学报, 2019, 35(18): 143-150.
[5] Xu Y, Imou K, Kaizu Y, et al. Two-stage approach for detecting slightly overlapping strawberries using HOG descriptor [J]. Biosystems Engineering, 2013, 115(2): 144-153.
[6] Arivazhagan S, Shebiah R N, Nidhyanandhan S S, et al. Fruit recognition using color and texture features [J]. Journal of Emerging Trends in Computing and Information Sciences, 2010, 2(1): 90-94.
[7] 陶华伟, 赵力, 奚吉, 等.基于颜色及纹理特征的果蔬种类识别方法[J]. 农业工程学报, 2014, 30(16): 305-311.
[8] 傅隆生, 冯亚利, Elkamil T, 等. 基于卷积神经网络的田间多簇猕猴桃图像识别方法[J]. 农业工程学报, 2018, 34(2): 205-211.
[9] 张小花, 马瑞峻, 吴卓葵, 等. 基于机器视觉的果园成熟柑橘快速识别及产量预估研究[J]. 广东农业科学, 2019, 46(7): 156-161.
[10] 廖崴, 郑立华, 李民赞, 等. 基于随机森林算法的自然光照条件下绿色苹果识别[J]. 农业机械学报, 2017, 48(S1): 86-91.
[11] Kurtulmus F, Lee W S, Vardar A. Green citrus detection using “eigenfruit”, color and circular Gabor texture features under natural outdoor conditions [J]. Computers and Electronics in Agriculture, 2011, 78: 140-149.
[12] Sengupta S, Lee W S. Identification and determination of the number of immature green citrus fruit in a canopy under different ambient light conditions [J]. Biosystems Engineering, 2014, 117: 51-61.
[13] Zhao C Y, Lee W S, He D J. Immature green citrus detection based on color feature and sum of absolute transformed difference (SATD) using color images in the citrus grove [J]. Computers and Electronics in Agriculture, 2016, 124: 243-253.
[14] Gan H, Lee W S, Alchanatis V, et al. Immature green citrus fruit detection using color and thermal images [J]. Computers and Electronics in Agriculture, 2018, 152: 117-125.
[15] 熊俊涛, 刘振, 汤林越, 等. 自然环境下绿色柑橘视觉检测技术研究[J]. 农业机械学报, 2018, 49(4): 45-52.
[16] 吕石磊, 卢思华, 李震, 等. 基于改进YOLOv3-LITE轻量级神经网络的柑橘识别方法[J]. 农业工程学报, 2019, 35(17): 205-214.
[17] Wu X, Sahoo D, Steven C H H. Recent advances in deep learning for object detection [J]. Neurocomputing, 2020, 396: 39-64.
[18] Redmon J, Farhadi A. YOLO v3: An incremental improvement [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2018.
[19] Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2016: 779-788.
[20] Redmon J, Farhadi A. YOLO9000: better, faster,stronger [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2017: 6517-6525.
[21] Gao H, Zhang L. Densely connected convolutional networks [C]. IEEE conference on Computer Vision and Pattern Recognition, 2017: 1-28.
[22] He K, Zhang X, Ren S, et al. Deep residual learning for image recognition [C]. IEEE conference on Computer Vision and Pattern Recognition, 2016: 770-778.

基于改进YOLOV3的自然环境下绿色柑橘的识别算法^*

Identification of green citrus based on improved YOLOV3 in natural environment

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

编辑推荐

Metrics

本文评价

[1]	何颖, 陈丁号, 彭琳. 基于改进YOLOv5模型的经济林木虫害目标检测算法研究[J]. 中国农机化学报, 2022, 43(4): 106-115.
[2]	李小敏;张日红;陈天赐;侯炳法;张权;李雄;. 基于深度学习的林下落果识别方法与试验[J]. 中国农机化学报, 2021, 42(9): 202-208.
[3]	项新建;周跃琪;姚佳娜;宋晓敏;郑永平;. 野外农用视频监控运动目标检测算法研究及系统开发[J]. 中国农机化学报, 2021, 42(5): 166-174.
[4]	成伟;张文爱;冯青春;张万豪;. 基于改进YOLOv3的温室番茄果实识别估产方法[J]. 中国农机化学报, 2021, 42(4): 176-182.
[5]	闫建伟;张乐伟;赵源;张富贵;. 改进RetinaNet的刺梨果实图像识别[J]. 中国农机化学报, 2021, 42(3): 78-83.
[6]	李震;颜少华;洪添胜;吕石磊;. 基于深度相机的山地果园运输车避障系统设计[J]. 中国农机化学报, 2021, 42(2): 115-120.
[7]	王林柏, 刘景艳, 周玉宏, 张君, 李兴旺, 范晓飞. 　基于分水岭算法结合卷积神经网络的玉米种子质量检测[J]. 中国农机化学报, 2021, 42(12): 168-174.
[8]	王林柏, 张博, 姚竟发, 杨志辉, 张君, 范晓飞. 基于卷积神经网络马铃薯叶片病害识别和病斑检测^*[J]. 中国农机化学报, 2021, 42(11): 122-129.