Detection and counting of wheat ears based on YOLOv3 network

doi:10.14088/j.cnki.issn0439-8114.2021.02.033

Abstract

Abstract: Wheat ear detection and counting are very important for wheat yield estimation and breeding. However, traditional wheat ear quantity statistics are based on manual statistical methods or remote sensing prediction methods, which are inefficient and not accurate. In order to solve the above problems, this paper proposes a wheat detection method based on YOLOv3 deep neural network. The results showed that YOLOv3 has achieved an average accuracy mAP value of 67.81% and an accuracy rate of 93% for wheat ear counting on three common wheat varieties, which can quickly and efficiently detect wheat ears in a specific label box.

Key words: wheat ear, YOLOv3 network, wheat ear detection, wheat ear counting

CLC Number:

TP391.4

HAO Wang-li, WEI Pei-yan, HAN Meng, ZHANG Li, XI Rui-ze. Detection and counting of wheat ears based on YOLOv3 network[J]. HUBEI AGRICULTURAL SCIENCES, 2021, 60(2): 158-160.

References

[1] 甘肃省科学院生物所. 小麦测产新方法[P].中国专利:CN1031424A.
[2] 侯英鱼, 王健林, 毛留喜,等. 美国玉米和小麦产量动态预测遥感模型[J]. 生态学杂志, 2009, 28(10): 2141-2146.
[3] 朱秀红, 李秀珍, 姚文军,等. 基于SPSS的日照市小麦产量年景预测模型[J]. 中国农学通报, 2010, 26(2): 295-297.
[4] 吕行军. 基于图像处理技术的麦穗识别研究[D]. 河北保定:河北农业大学, 2012.
[5] BENGIO Y, COURVILLE A, VINCENT P.Representation learning: A review and new perspectives[J]. IEEE transactions on pattern analysis &machine intelligence, 2012, 35(8): 1798-1828.
[6] 周飞燕, 金林鹏, 董军. 卷积神经网络研究综述[J]. 计算机学报, 2017, 40(6): 1229-1251.
[7] GIRSHICK R, DONAHUUE J, DARRELL T, et al.Rich feature hierarchies for accurate object detection and semantic segmentation[A]//proceeding of the IEEE Conference on Computer Vision and Pattern Recognition, 2014: 580-587.
[8] HE K, ZHANGX, RENS, et al.Spatisl pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions Pattern Anal Mach Intell, 2015(9): 1904-1916.
[9] GIRSHICK R.Fast R-CNN[A].IEEE International Conference on Computer Vision, 2015:1440-1448.
[10] REN S, HE K, GIRSHICK R, et al.Faster R-CNN: Towards real-time object detection with region proposal networks[C]//International Conference on Neural Information Processing Systems. Cambridge: MIT, 2015: 91-99.
[11] LI Y, HE K, SUN J.R-FCN: Object detection via region based fully convolutional networks[R]. arXiv, 2016.
[12] REDMON J, DIVVALA S, GIRSHICK R, et al.You only look once: Unified, real-time object detection[A].IEEE Conference on Computer Vision and Pattern Recognition[C]. 2016: 779-788.
[13] LIU W, ANGUELOV D, ERHAN D, et al.SSD: Single shot multibox detector[R]. arXiv, 2016.
[14] REDMON J, FARHADI A.YOLO9000: Better, faster, stronger[R]. arXiv, 2016.
[15] REDMON J, FARHADI A.YOLOv3: An Incremental Improvement[R]. arXiv, 2018.
[16] 郭龙源, 童光红, 段厚裕, 等. 基于YOLOv3的磁瓦表面缺陷检测算法[J]. 成都工业学院学报, 2019, 22(3): 25-30.
[18] 王毅恒, 许德章. 基于YOLOv3算法的农场环境下奶牛目标识别[J]. 广东石油化工学院学报, 2019, 29(4): 31-35.