JIA-2022-0425 MRUNet:一种复杂环境下昆虫小目标的两阶段分割模型

doi:10.1016/j.jia.2022.09.004

Journal of Integrative Agriculture ›› 2023, Vol. 22 ›› Issue (4): 1117-1130.DOI: 10.1016/j.jia.2022.09.004

JIA-2022-0425 MRUNet:一种复杂环境下昆虫小目标的两阶段分割模型

收稿日期:2022-04-02 接受日期:2022-07-11 出版日期:2023-04-20 发布日期:2022-07-11

MRUNet: A two-stage segmentation model for small insect targets in complex environments

WANG Fu-kuan^{1, 2*}, HUANG Yi-qi^1*, HUANG Zhao-cheng³, SHEN Hao^{1, 2}, HUANG Cong², QIAO Xi^{1, 2#}, QIAN Wan-qiang^2#

¹College of Mechanical Engineering, Guangxi University, Nanning 530004, P.R.China
²Lingnan Guangdong Laboratory of Modern Agriculture, Genome Analysis Laboratory of the Ministry of Agriculture and Rural Affairs, Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences, Shenzhen 518120, P.R.China
³Wuzhou Product Quality Inspection Institute, Wuzhou 543000, P.R.China

Received:2022-04-02 Accepted:2022-07-11 Online:2023-04-20 Published:2022-07-11
About author: WANG Fu-kuan, E-mail: wfukuan@163.com; HUANG Yi-qi, E-mail: hyqgxu@163.com; #Correspondence QIAN Wan-qiang, E-mail: qianwanqiang@caas.cn; QIAO Xi, E-mail: qiaoxi@ caas.cn * These authors contributed equally to this study.
Supported by:
The study in this paper was supported by the National Key Research and Development Program of China
(2021YFD1400100, 2021YFD1400101 and 2021YFD1400102), the Guangxi Natural Science Foundation of China (2021JJA130221), and the Shenzhen Science and Technology Program, China (KQTD20180411143628272).

摘要/Abstract

摘要：

田间害虫在线自动识别是农田害虫防治的重要辅助手段。在实际应用中，由于目标小、物种相似性高、背景复杂等因素，昆虫在线识别系统往往无法对目标进行准确定位和识别。为方便昆虫幼虫的识别，本研究提出了一种两阶段分割方法MRUNet。在结构上，MRUNet借鉴了Mask R-CNN在进行语义分割之前先进行目标检测的做法，并且使用改进的轻量级UNet进行语义分割。为了可靠地评价所提出模型的分割结果，除了常用的语义分割评价指标外，本研究还引入了统计方法来衡量模型在样本间性能的稳定性。实验结果表明，所提出的两阶段图像分割策略能有效处理复杂背景下的小目标。与现有的语义分割方法相比，MRUNet在相同条件下具有更好的稳定性和细节处理能力。本研究可以为昆虫幼虫识别提供高质量的图像数据。

Abstract:

Online automated identification of farmland pests is an important auxiliary means of pest control. In practical applications, the online insect identification system is often unable to locate and identify the target pest accurately due to factors such as small target size, high similarity between species and complex backgrounds. To facilitate the identification of insect larvae, a two-stage segmentation method, MRUNet was proposed in this study. Structurally, MRUNet borrows the practice of object detection before semantic segmentation from Mask R-CNN and then uses an improved lightweight UNet to perform the semantic segmentation. To reliably evaluate the segmentation results of the models, statistical methods were introduced to measure the stability of the performance of the models among samples in addition to the evaluation indicators commonly used for semantic segmentation. The experimental results showed that this two-stage image segmentation strategy is effective in dealing with small targets in complex backgrounds. Compared with existing state-of-the-art semantic segmentation methods, MRUNet shows better stability and detail processing ability under the same conditions. This study provides a reliable reference for the automated identification of insect larvae.

Key words: insect larvae , deep learning , semantic segmentation , object detection

WANG Fu-kuan, HUANG Yi-qi, HUANG Zhao-cheng, SHEN Hao, HUANG Cong, QIAO Xi, QIAN Wan-qiang. JIA-2022-0425 MRUNet:一种复杂环境下昆虫小目标的两阶段分割模型[J]. Journal of Integrative Agriculture, 2023, 22(4): 1117-1130.

WANG Fu-kuan, HUANG Yi-qi, HUANG Zhao-cheng, SHEN Hao, HUANG Cong, QIAO Xi, QIAN Wan-qiang.

MRUNet: A two-stage segmentation model for small insect targets in complex environments [J]. Journal of Integrative Agriculture, 2023, 22(4): 1117-1130.

参考文献

Ali M, Gilani S O, Waris A, Zafar K, Jamil M. 2020. Brain tumour image segmentation using deep networks. Ieee Access, 8, 153589–153598.
Alves A N, Souza W S R, Borges D L. 2020. Cotton pests classification in field-based images using deep residual networks. Computers and Electronics in Agriculture, 174, 105488.
Bai R F, Jiang S, Sun H J, Yang Y F, Li G J. 2021. Deep neural network-based semantic segmentation of microvascular decompression images. Sensors, 21, 1167.
Bochkovskiy A, Wang C Y, Liao H Y M. 2020. YOLOv4: Optimal speed and accuracy of object detection. ArXiv, doi: 10.48550/arXiv.2004.10934.
Canny J. 1986. A computational approach to edge detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 8, 679–698.
Chen L C E, Zhu Y K, Papandreou G, Schroff F, Adam H. 2018. Encoder-Decoder with atrous separable convolution for semantic image segmentation. In: The 15th European Conference on Computer Vision (ECCV). Springer, Cham, Munich, Germany. pp. 833–851.
Dadashzadeh A, Targhi A T, Tahmasbi M, Mirmehdi M. 2019. HGR-Net: A fusion network for hand gesture segmentation and recognition. Iet Computer Vision, 13, 700–707.
Early R, Gonzalez-Moreno P, Murphy S T, Day R. 2018. Forecasting the global extent of invasion of the cereal pest Spodoptera frugiperda, the fall armyworm. Neobiota, 40, 25–50.
He K M, Gkioxari G, Dollar P, Girshick R. 2017. Mask R-CNN. In: The 16th IEEE International Conference on Computer Vision (ICCV). IEEE, Venice, Italy. pp. 2980–2988.
Li F, Xiong Y. 2018. Automatic identification of butterfly species based on HoMSC and GLCMoIB. Visual Computer, 34, 1525–1533.
Li K S, Xiong L, Zhang D B, Liang Z P, Xue Y. 2017. The research of disease spots extraction based on evolutionary algorithm. Journal of Optimization, 2017, 4093973.
Li Q W, Jia W K, Sun M L, Hou S J, Zheng Y J. 2021. A novel green apple segmentation algorithm based on ensemble U-Net under complex orchard environment. Computers and Electronics in Agriculture, 180, 105900.
Li X M, Dai B S, Sun H M, Li W N. 2019. Corn classification system based on computer vision. Symmetry (Basel), 11, 591.
Long J, Shelhamer E, Darrell T. 2015. Fully convolutional networks for semantic segmentation. In: The 2015 IEEEConference on Computer Vision and Pattern Recognition (CVPR). IEEE, Boston, USA. pp. 3431–3440.
Lu S H, Ye S J. 2020. Using an image segmentation and support vector machine method for identifying two locust species and instars. Journal of Integrative Agriculture, 19, 1301–1313.
Nascimento D A, Anunciacao R M, Arnhold A, Ferraz A C, Santos A D, Zanuncio J C. 2016. Expert system for identification of economically important insect pests in commercial teak plantations. Computers and Electronics in Agriculture, 121, 368–373.
Otsu N. 1979. A threshold selection method from gray-level histograms. IEEE Transactions on Systems, Man, and Cybernetics, 9, 62–66.
Pan Q, Gao M F, Wu P B, Yan J W, Li S L. 2021. A deep-learning-based approach for wheat yellow rust disease recognition from unmanned aerial vehicle images. Sensors, 21, 6540.
Ronneberger O, Fischer P, Brox T. 2015. U-Net: Convolutional networks for biomedical image segmentation. In: The 18th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). Springer, Cham, Munich, GERMANY. pp. 234–241.
Rother C, Kolmogorov V, Blake A. 2004. “GrabCut” - Interactive foreground extraction using iterated graph cuts. Acm Transactions on Graphics, 23, 309–314.
Saleem S, Amin J, Sharif M, Anjum M A, Iqbal M, Wang S H. 2022. A deep network designed for segmentation and classification of leukemia using fusion of the transfer learning models. Complex & Intelligent Systems, 8, 3105–3120.
Sandler M, Howard A, Zhu M L, Zhmoginov A, Chen L C. 2018. MobileNetV2: Inverted residuals and linear bottlenecks. In: The 31st IEEE/CVFConference on Computer Vision and Pattern Recognition (CVPR). IEEE, Salt Lake City, UT. pp. 4510–4520.
Tang H, Wang B, Chen X. 2020. Deep learning techniques for automatic butterfly segmentation in ecological images. Computers and Electronics in Agriculture, 178, 105739.
Tenorio G L, Martins F F, Carvalho T M, Leite A C, Figueiredo K, Vellasco M, Caarls W. 2019. Comparative study of computer vision models for insect pest identification in complex backgrounds. In: The 12th International Conference on the Developments in Systems Engineering (DeSE). IEEE, Kazan, Russia. pp. 551–556.
Tong M S, Li W C, Ren X Y, Yu X H, Lin W Y. 2020. Weakly-supervised semantic segmentation with regional location cutting and dynamic credible regions correction. Ieee Access, 8, 204378–204388.
Vincent L, Soille P. 1991. Watersheds in digital spaces: An efficient algorithm based on immersion simulations. IEEE Transactions on Pattern Analysis and Machine Intelligence, 13, 583–598.
Wang C S, Du P F, Wu H R, Li J X, Zhao C J, Zhu H J. 2021. A cucumber leaf disease severity classification method based on the fusion of DeepLabV3+ and U-Net. Computers and Electronics in Agriculture, 189, 106373.
Wu Z C, Yang R Z, Gao F F, Wang W Q, Fu L S, Li R. 2021. Segmentation of abnormal leaves of hydroponic lettuce based on DeepLabV3+for robotic sorting. Computers and Electronics in Agriculture, 190, 106443.
Xia C L, Chon T S, Ren Z M, Lee J M. 2015. Automatic identification and counting of small size pests in greenhouse conditions with low computational cost. Ecological Informatics, 29, 139–146.
Xu L L, Li Y, Xu J Z, Guo L L. 2020. Two-level attention and score consistency network for plant segmentation. Computers and Electronics in Agriculture, 170, 105281.
Yu Y, Zhang K L, Yang L, Zhang D X. 2019. Fruit detection for strawberry harvesting robot in non-structural environment based on Mask-RCNN. Computers and Electronics in Agriculture, 163, 104846.
Zhang D J, Pan Y Z, Zhang J S, Hu T G, Zhao J H, Li N, Chen Q. 2020. A generalized approach based on convolutional neural networks for large area cropland mapping at very high resolution. Remote Sensing of Environment, 247, 111912.
Zhao H S, Shi J P, Qi X J, Wang X G, Jia J Y. 2017. Pyramid scene parsing network. In: The 30th IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, Honolulu, HI. pp. 6230–6239.

JIA-2022-0425 MRUNet:一种复杂环境下昆虫小目标的两阶段分割模型

MRUNet: A two-stage segmentation model for small insect targets in complex environments

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

编辑推荐

Metrics