农业复杂环境下尺度自适应小目标识别算法——以蜜蜂为研究对象

doi:10.12133/j.smartag.SA202203003

Smart Agriculture ›› 2022, Vol. 4 ›› Issue (1): 140-149.doi: 10.12133/j.smartag.SA202203003

• 信息处理与决策 • 上一篇

农业复杂环境下尺度自适应小目标识别算法——以蜜蜂为研究对象

郭秀明(), 诸叶平, 李世娟, 张杰, 吕纯阳, 刘升平()

中国农业科学院农业信息研究所/农业农村部农业信息服务技术重点实验室，北京 100081

收稿日期:2021-12-31 出版日期:2022-03-30
基金资助:
中央级公益性科研院所基本科研业务费专项(2021JKY038);河北省重点研发计划项目子课题(19227407D);国家蜂产业技术体系专项(CARS-44)
作者简介:郭秀明（1981－），女，博士，副研究员，研究方向为农业智能感知。E-mail：guoxiuming@caas.cn。
通信作者: 刘升平（1981－），男，博士，研究员，研究方向为农产品质量安全控制、作物模拟模型技术等。E-mail：liushengping@caas.cn

Scale Adaptive Small Objects Detection Method in Complex Agricultural Environment: Taking Bees as Research Object

GUO Xiuming(), ZHU Yeping, LI Shijuan, ZHANG Jie, LYU Chunyang, LIU Shengping()

Agricultural Information Institute, Chinese Academy of Agricultural Sciences/Key Laboratory of Agri-information Service Technology, Ministry of Agriculture and Rural Affairs, Beijing 100081, China

Received:2021-12-31 Online:2022-03-30
corresponding author: LIU Shengping, E-mail：liushengping@caas.cn
About author:GUO Xiuming, E-mail：guoxiuming@caas.cn
Supported by:
Special Fund for Basic Scientific Research Business of Central-level Public Welfare Scientific Research Institutes(2021JKY038);Key Research and Development Plan Project of Hebei Province (19227407D);National Bee Industry Technology System Project (CARS-44)

摘要/Abstract

摘要：

农业生产环境中的目标识别对象常具有分布密集、体积小、密度大的特点，加之农田环境光照多变、背景复杂，导致已有目标检测模型无法取得令人满意的效果。本研究以提高小目标的识别性能为目标，以蜜蜂识别为例，提出了一种农业复杂环境下尺度自适应小目标识别算法。算法克服了复杂多变的背景环境的影响及目标体积较小导致的特征提取困难，实现目标尺度无关的小目标识别。首先将原图拆分为一些较小尺寸的子图以提高目标尺度，将已标注的目标分配到拆分后的子图中，形成新的数据集，然后采用迁移学习的方法重新训练并生成新的目标识别模型。在模型的使用中，为使子图识别结果能正常还原，拆分的子图之间需具有一定的重叠率。收集所有子图的目标识别结果，采用非极大抑制（Non-Maximum Suppression，NMS）去除由于模型本身产生的冗余框，提出一种交小比非极大抑制（Intersection over Small NMS，IOS-NMS）进一步去除子图重叠区域中的冗余框。在子图像素尺寸分别为300×300、500×500和700×700，子图重叠率分别为0.2和0.05的情况下进行验证试验，结果表明：采用SSD（Single Shot MultiBox Detector）作为框架中的目标检测模型，新提出的尺度自适应算法的召回率和精度普遍高于SSD模型，最高分别提高了3.8%和2.6%，较原尺度的YOLOv3模型也有一定的提升。为进一步验证算法在复杂背景中小目标识别的优越性，从网上爬取了不同尺度、不同场景的农田复杂环境下的蜜蜂图像，并采用本算法和SSD模型进行了对比测试，结果表明：本算法能提高目标识别性能，具有较强的尺度适应性和泛化性。由于本算法对于单张图像需要多次向前推理，时效性不高，不适用于边缘计算。

关键词: 目标检测, 机器视觉, 小目标, 农业环境, 蜜蜂, SSD, YOLOv3

Abstract:

Objects in farmlands often have characteristic of small volume and high density with variable light and complex background, and the available object detection models could not get satisfactory recognition results. Taking bees as research objects, a method that could overcome the influence from the complex backgrounds, the difficulty in small object feature extraction was proposed, and a detection algorithm was created for small objects irrelevant to image size. Firstly, the original image was split into some smaller sub-images to increase the object scale, and the marked objects were assigned to the sub-images to produce a new dataset. Then, the model was trained again using transfer learning to get a new object detection model. A certain overlap rate was set between two adjacent sub-images in order to restore the objects. The objects from each sub-image was collected and then non-maximum suppression (NMS) was performed to delete the redundant detection boxes caused by the network, an improved NMS named intersection over small NMS (IOS-NMS) was then proposed to delete the redundant boxes caused by the overlap between adjacent sub-images. Validation tests were performed when sub-image size was set was 300×300, 500×500 and 700×700, the overlap rate was set as 0.2 and 0.05 respectively, and the results showed that when using single shot multibox detector (SSD) as the object detection model, the recall rate and precision was generally higher than that of SSD with the maximum difference 3.8% and 2.6%, respectively. In order to further verify the algorithm in small target recognition with complex background, three bee images with different scales and different scenarios were obtained from internet and test experiments were conducted using the new proposed algorithm and SSD. The results showed that the proposed algorithm could improve the performance of target detection and had strong scale adaptability and generalization. Besides, the new algorithm required multiple forward reasoning for a single image, so it was not time-efficient and was not suitable for edge calculation.

Key words: object detection, machine vision, small object, farmland, bee, single shot multibox detector, YOLOv3

中图分类号:

郭秀明, 诸叶平, 李世娟, 张杰, 吕纯阳, 刘升平. 农业复杂环境下尺度自适应小目标识别算法——以蜜蜂为研究对象[J]. 智慧农业(中英文), 2022, 4(1): 140-149.

GUO Xiuming, ZHU Yeping, LI Shijuan, ZHANG Jie, LYU Chunyang, LIU Shengping. Scale Adaptive Small Objects Detection Method in Complex Agricultural Environment: Taking Bees as Research Object[J]. Smart Agriculture, 2022, 4(1): 140-149.

参考文献

1	NAUATA N, HU H, ZHOU G T, et al. Structured label inference for visual understanding[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 42(5): 1257-1271.
2	黄凯奇, 任伟强, 谭铁牛. 图像物体分类与检测算法综述[J]. 计算机学报, 2014, 36(12):1-18.
2	HUANG K, REN W, TAN T. A review on image object classification and detection[J]. Chinese Journal of Computers, 2014, 36(12):1-18.
3	ZOU Z, SHI Z, GUO Y, et al. Object detection in 20 years: A survey[J/OL]. arXiv: 1905.05055v2 [cs.CV], 2019.
4	李科岑, 王晓强, 林浩, 等. 深度学习中的单阶段小目标检测方法综述[J]. 计算机科学与探索, 2022, 16(1): 41-58.
4	LI K, WANG X, LIN H, et al. Survey of one stage small object detection methods in deep learning[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(1): 41-58.
5	BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOv4: Optimal speed and accuracy of object detection[J/OL]. arXiv: , 2020.
6	REDMON J, FARHADI A. YOLOv3: An incremental improvement[J/OL]. arXiv:1804.02767 [cs.CV], 2018.
7	MAHTO P, GARG P, SETH P, et al. Refining YOLOv4 for vehicle detection[J]. International Journal of Advanced Research in Engineering and Technology, 2020, 11(5): 409-419.
8	ZHAI S, SHANG D, WANG S, et al. DF-SSD: An improved SSD object detection algorithm based on denseNet and feature fusion[J]. IEEE Access, 2020: 24344-24357.
9	奚琦, 张正道, 彭力. 基于改进密集网络与二次回归的小目标检测算法[J]. 计算机工程, 2021, 47(4): 241-247, 255.
9	XI Q, ZHANG Z, PENG L. Small object detection algorithm based on improved dense network and quadratic regression[J]. Computer Engineering, 2021, 47(4): 241-247, 255.
10	SHENZ Q, LIU Z, LI J G, et al. DSOD: Learning deeply supervised object detectors from scratch[C]// The 2017 IEEE International Conference on Computer Vision. Washington D. C., USA: IEEE Computer Society, 2017: 1937-1945.
11	李航, 朱明. 基于深度卷积神经网络的小目标检测算法[J]. 计算机工程与科学, 2020, 42(4): 649-657.
11	LI H, ZHU M. A small object detection algorithm based on deep convolutional neural network[J]. Computer Engineering & Science, 2020, 42(4): 649-657.
12	周慧, 严凤龙, 褚娜, 等. 一种改进复杂场景下小目标检测模型的方法[J/OL]. 计算机工程与应用: 1-8. [2021-10-04]. .
12	ZHOU H, YAN F, ZHU N, et al. An approach to improve the detection model for small object in complex scenes[J/OL]. Computer Engineering and Applications:1-8. [2021-10-04]. .
13	ESTER M, KRIEGEL H P, SANDER J, et al. A density-based algorithm for discovering clusters in large spatial databases with noise[C]// The Second International Conference on Knowledge Discovery and Data Mining. Portland, Oregon, USA: AAAI, 1996: 226-231.
14	李云红, 张轩, 李传真, 等. 融合DBSCAN的改进YOLOv3目标检测算法[J/OL]. 计算机工程与应用: 1-12. [2021-10-04]. .
14	LI Y, ZHANG X, LI C, et al. Improved YOLOv3 target detection algorithm combined with DBSCAN[J/OL]. Computer Engineering and Applications: 1-12. [2021-10-04]. .
15	REZATOFIGHI H, TSOI N, GWAK J Y, et al. Generalized intersection over union: A metric and a loss for bounding box regression[C]// IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, New York, USA: IEEE, 2019: 658-666.
16	YANG Y, LIAO Y, CHENG L, et al. Remote sensing image aircraft target detection based on GIoU-YOLOv3[C]// 2021 6th International Conference on Intelligent Computing and Signal Processing. Piscataway, New York, USA: IEEE, 2021: 474-478.
17	ZHENG Z, ZHAO H, LIU W, et al. Distance-IoU loss: Faster and better learning for bounding box regression[C]// The 34th AAAI Conference on Artificial Intelligence, the 32nd Innovative Applications of Artificial Intelligence Conference, the 10th AAAI Symposium on Educational Advances in Artificial Intelligence. Piscataway, New York, USA: AAAI, 2020: 12993-13000.
18	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017, (99): 2999-3007.
19	张炳力, 秦浩然, 江尚, 等. 基于RetinaNet 及优化损失函数的夜间车辆检测方法[J]. 汽车工程, 2021, 43(8): 1195-1202.
19	ZHANG B, QIN H, JIANG S, et al. A method of vehicle detection at night based on RetinaNet and optimized loss functions[J]. Automotive Engineering, 2021, 43(8): 1195-1202.
20	郑秋梅, 王璐璐, 王风华. 基于改进卷积神经网络的交通场景小目标检测[J]. 计算机工程, 2020, 46(6): 26-33.
20	ZHENG Q, WANG L, WANG F. Small object detection in traffic scene based on improved convolutional neural network[J]. Computer Engineering, 2020, 46(6): 26-33.
21	REN S, HE K, GIRSHICK R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
22	LIU W, ANGUELOV D, ERHAN D, et al. SSD: Single shot multibox detector[C]// In European Conference on Computer Vision. Cham, Switzerland: Springer: 2016.
23	Neubeck A, Gool L J V. Efficient non-maximum suppression[C]// International Conference on Pattern Recognition. Piscataway, New York, USA: IEEE Computer Society, 2006: 848-855.
24	李景琳, 姜晶菲, 窦勇, 等. 基于Soft-NMS的候选框去冗余加速器设计[J]. 计算机工程与科学, 2021, 43(4): 586-593.
24	LI J, JIANG J, DOU Y, et al. A redundacy-reduced candidate box accelerator based on soft-non-maximum suppression[J]. Computer Engineering & Science, 2021, 43(4): 586-593.
25	张长伦, 张翠文, 王恒友, 等. 基于注意力机制的NMS在目标检测中的研究[J]. 电子测量技术, 2021, 44(19): 82-88.
25	ZHANG C, ZHANG C, WANG H, et al. Research on non-maximum suppression based on attention mechanism in object detection[J]. Electronic Measurement Technology, 2021, 44(19): 82-88.

[1]	吴小燕, 郭威, 朱轶萍, 朱华吉, 吴华瑞. 基于改进YOLOv8s的大田甘蓝移栽状态检测算法[J]. 智慧农业(中英文), 2024, 6(2): 107-117.
[2]	毛永文, 韩俊英, 刘成忠. 基于机器视觉的胡麻种子自动化考种方法[J]. 智慧农业(中英文), 2024, 6(1): 135-146.
[3]	李政凯, 于嘉辉, 潘时佳, 贾泽丰, 牛子杰. 冬季猕猴桃树单木骨架提取与冠层生长预测方法[J]. 智慧农业(中英文), 2023, 5(4): 92-104.
[4]	李阳德, 马晓慧, 王骥. 基于轻量级MobileNet V3-YOLOv4的生长期菠萝成熟度分析[J]. 智慧农业(中英文), 2023, 5(2): 35-44.
[5]	商枫楠, 周学成, 梁英凯, 肖明玮, 陈桥, 罗陈迪. 基于改进YOLOX的自然环境中火龙果检测方法[J]. 智慧农业(中英文), 2022, 4(3): 120-131.
[6]	何锐敏, 郑可锋, 尉钦洋, 张小斌, 张俊, 朱怡航, 赵懿滢, 顾清. 基于改进Mask R-CNN模型的工厂化养蚕蚕体识别与计数[J]. 智慧农业(中英文), 2022, 4(2): 163-173.
[7]	郭志明, 王郡艺, 宋烨, 邹小波, 蔡健荣. 果蔬品质劣变传感检测与监测技术研究进展[J]. 智慧农业(中英文), 2021, 3(4): 14-28.
[8]	龙洁花, 郭文忠, 林森, 文朝武, 张宇, 赵春江. 改进YOLOv4的温室环境下草莓生育期识别方法[J]. 智慧农业(中英文), 2021, 3(4): 99-110.
[9]	郭威, 吴华瑞, 朱华吉. 设施温室影像采集与环境监测机器人系统设计及应用[J]. 智慧农业(中英文), 2020, 2(3): 48-60.
[10]	韩书庆, 张晶, 程国栋, 彭英琦, 张建华, 吴建寨. 奶牛跛行自动识别技术研究现状与挑战[J]. 智慧农业(中英文), 2020, 2(3): 21-36.
[11]	夏雪, 孙琦鑫, 侍啸, 柴秀娟. 基于轻量级无锚点深度卷积神经网络的树上苹果检测模型[J]. 智慧农业(中英文), 2020, 2(1): 99-110.
[12]	周成全, 叶宏宝, 俞国红, 胡俊, 徐志福. 基于机器视觉与深度学习的西兰花表型快速提取方法研究[J]. 智慧农业(中英文), 2020, 2(1): 121-132.
[13]	周超, 徐大明, 吝凯, 陈澜, 张松, 孙传恒, 杨信廷. 基于近红外机器视觉的鱼类摄食强度评估方法研究[J]. 智慧农业(中英文), 2019, 1(1): 76-84.

农业复杂环境下尺度自适应小目标识别算法——以蜜蜂为研究对象

Scale Adaptive Small Objects Detection Method in Complex Agricultural Environment: Taking Bees as Research Object

在线阅读

知网下载

本地下载

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 13

编辑推荐

Metrics

本文评价