AI Chat Paper
Note: Please note that the following content is generated by AMiner AI. SciOpen does not take any responsibility related to this content.
{{lang === 'zh_CN' ? '文章概述' : 'Summary'}}
{{lang === 'en_US' ? '中' : 'Eng'}}
Chat more with AI
Powered by AMiner. Clicking this button will take you away from SciOpen.
Home Plant Phenomics Article
Article Link
Cite
EndNote(RIS) BibTeX
Collect
Submit Manuscript
Show Outline
Outline
Show full outline
Hide outline
Outline
Show full outline
Hide outline
Research Article | Open Access

Lightweight Deep Learning Models for High-Precision Rice Seedling Segmentation from UAV-Based Multispectral Images

Panli Zhang1, Xiaobo Sun1( ), Donghui Zhang2, Yuechao Yang3, Zhenhua Wang4
College of Engineering, Northeast Agricultural University, Harbin 150030, China
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China
National Key Laboratory of Remote Sensing Information and Imagery Analyzing Technology, Beijing Research Institute of Uranium Geology, Beijing 100029, China
College of Agriculture, Northeast Agricultural University, Harbin 150030, China
Show Author Information

Abstract

Accurate segmentation and detection of rice seedlings is essential for precision agriculture and high-yield cultivation. However, current methods suffer from high computational complexity and poor robustness to different rice varieties and densities. This article proposes 2 lightweight neural network architectures, LW-Segnet and LW-Unet, for high-precision rice seedling segmentation. The networks adopt an encoder–decoder structure with hybrid lightweight convolutions and spatial pyramid dilated convolutions, achieving accurate segmentation while reducing model parameters. Multispectral imagery acquired by unmanned aerial vehicle (UAV) was used to train and test the models covering 3 rice varieties and different planting densities. Experimental results demonstrate that the proposed LW-Segnet and LW-Unet models achieve higher F1-scores and intersection over union values for seedling detection and row segmentation across varieties, indicating improved segmentation accuracy. Furthermore, the models exhibit stable performance when handling different varieties and densities, showing strong robustness. In terms of efficiency, the networks have lower graphics processing unit memory usage, complexity, and parameters but faster inference speeds, reflecting higher computational efficiency. In particular, the fast speed of LW-Unet indicates potential for real-time applications. The study presents lightweight yet effective neural network architectures for agricultural tasks. By handling multiple rice varieties and densities with high accuracy, efficiency, and robustness, the models show promise for use in edge devices and UAVs to assist precision farming and crop management. The findings provide valuable insights into designing lightweight deep learning models to tackle complex agricultural problems.

References

1

Wang J, Sun X, Xu Y, Wang Q, Tang H, Zhou W. The effect of harvest date on yield loss of long and short-grain rice cultivars (Oryza sativa L.) in Northeast China. Eur J Agron. 2021;131(8):126382.
Crossref Google Scholar
2

Lourenço P, Godinho S, Sousa A, Gonçalves AC. Estimating tree aboveground biomass using multispectral satellite-based data in Mediterranean agroforestry system using random forest algorithm. Remote Sens Appl: Soc Environ. 2021;23:100560.
Crossref Google Scholar
3

Yu Y, Bao Y, Wang J, Chu H, Zhao N, He Y, Liu Y. Crop row segmentation and detection in Paddy fields based on treble-classification Otsu and double-dimensional clustering method. Remote Sens. 2021;13(5):901.
Crossref Google Scholar
4

Wu J, Yang G, Yang X, Xu B, Han L, Zhu Y. Automatic counting of in situ rice seedlings from UAV images based on a deep fully convolutional neural network. Remote Sens. 2019;11(6):691–710.
Crossref Google Scholar
5

Varela S, Dhodda P, Hsu W, Prasad PVV, Id Y, Assefa Y, Peralta N, Griffin T, Sharda A, Ferguson A, et al. Early-season stand count determination in corn via integration of imagery from unmanned aerial systems (UAS) and supervised learning techniques. Remote Sens. 2018;10(2):10020343.
Crossref Google Scholar
6
Furukawa F, Maruyama K, Saito YK, Kaneko M. Corn height estimation using UAV for yield prediction and crop monitoring. In: Unmanned aerial vehicle: Applications in agriculture and environment. Cham (Switzerland): Springer International Publishing; 2020. p. 51–69.
Crossref
7

Comba L, Biglia A, Aimonino DR, Tortia C, Gay P. Leaf area index evaluation in vineyards using 3D point clouds from UAV imagery. Precis Agric. 2020;21:881–896.
Crossref Google Scholar
8

Mza B, Jz B, Kas C, Nrk C. Estimation of maize yield and effects of variable-rate nitrogen application using UAV-based RGB imagery. Biosyst Eng. 2020;189:24–35.
Crossref Google Scholar
9

Jibo Y, Feng H, Li Z, Zhou C, Xu K. Mapping winter-wheat biomass and grain yield based on a crop model and UAV remote sensing. Int J Remote Sens. 2020;42:1–25.
Crossref Google Scholar
10

Zhong L, Hu L, Zhou H. Deep learning based multi-temporal crop classification. Remote Sens Environ. 2019;221:430–443.
Crossref Google Scholar
11
Badeka E, Vrochidou E, Tziridis K, Nicolaou A, Kaburlasos VG. Navigation route mapping for harvesting robots in vineyards using UAV-based remote sensing. Paper presented at: 2020 IEEE 10th International Conference on Intelligent Systems (IS); 2020 Aug 28–30; Varna, Bulgaria.
Crossref
12
Jie YH, Liu ZA. Automatic track control method for multi-UAV based on embedded system. In: Advanced hybrid information processing. Changsha (China): Springer International Publishing; 2021. p. 399–408.
Crossref
13

Kitano BT, Mendes C, Geus AR, Oliveira HC, Souza JR. Corn plant counting using deep learning and UAV images. IEEE Geosci Remote Sens Lett. 2019;PP(99):1–5.
Crossref Google Scholar
14

Varela S, Dhodda PR, Hsu WH, Prasad PVV, Assefa Y, Peralta NR, Griffin T, Sharda A, Ferguson A, Ciampitti IA. Early-season stand count determination in corn via integration of imagery from unmanned aerial systems (UAS) and supervised learning techniques. Remote Sens. 2018;10(2):343.
Crossref Google Scholar
15

Su W, Jiang K, Yan A, Liu Z, Zhang M, Wang W. Monitoring of planted lines for breeding corn using UAV remote sensing image. Nongye Gongc Xuebao/Trans Chin Soc Agric Eng. 2018;34(5):92–98.
Google Scholar
16

Li D, Li B, Long S, Feng H, Xi T, Kang S, Wang JJBE. Rice seedling row detection based on morphological anchor points of rice stems. Biosyst Eng. 2023;226:71–85.
Crossref Google Scholar
17

Li D, Li B, Kang S, Feng H, Long S. E2CropDet: An efficient end-to-end solution to crop row detection. Expert Syst Appl. 2023;227:120345.
Crossref Google Scholar
18

Pang Y, Shi Y, Gao S, Jiang F, Sivakumar ANV, Thompson L, Luck J, Liu C. Improved crop row detection with deep neural network for early-season maize stand count in UAV imagery. Comput Electron Agric. 2020;178:Article 105766.
Crossref Google Scholar
19

Wu J, Guijun Y, Han L, Zhu Y. Automatic counting of in situ rice seedlings from UAV images based on a deep fully convolutional neural network. Remote Sens. 2019;11(6):11060691.
Crossref Google Scholar
20

Miyoshi G, Osco L, Junior J, Gonçalves D, Imai N, Tommaselli A, Honkavaara E, Gonçalves W, dos Santos M. A novel deep learning method to identify single tree species in UAV-based hyperspectral images. Remote Sens. 2020;12(8):12081294.
Crossref Google Scholar
21

Osco L, dos Santos de Arruda M, Gonçalves D, Dias A, Batistoti J, de Souza M, Gomes F, Ramos AP, Jorge L, Liesenberg V, et al. A CNN approach to simultaneously count plants and detect plantation-rows from UAV imagery. ISPRS J Photogrrammetry Remote Sens. 2021;174:1–17.
Crossref Google Scholar
22

Zhang E, Liu L, Huang L. An automated, generalized, deep-learning-based method for delineating the calving fronts of Greenland glaciers from multi-sensor remote sensing imagery. Remote Sens Environ. 2021;254:112265.
Crossref Google Scholar
23

Wang J, Sun X, Xu Y, Zhou W, Tang H, Wang Q. Timeliness harvesting loss of rice in cold region under different mechanical harvesting methods. Sustainability. 2021;13(11):6345.
Crossref Google Scholar
24

Zhuohuai G, Keyin C, Youchun D, Chongyou W. Visual navigation path extraction method in rice harvesting. Trans Chin Soc Agric Machinery. 2020;51(1):1.
Google Scholar
25

Kanagasingham S, Ekpanyapong M, Chaihan RJPA. Integrating machine vision-based row guidance with GPS and compass-based routing to achieve autonomous navigation for a rice field weeding robot. Precis Agric. 2020;21(4):831–855.
Crossref Google Scholar
26

Adhikari SP, Kim G, Kim HJIA. Deep neural network-based system for autonomous navigation in paddy field. IEEE Access. 2020;8(99):71272–71278.
Crossref Google Scholar
27

Zhang Q, Chen MS, Li BJC. A visual navigation algorithm for paddy field weeding robot based on image understanding. Comput Electron Agric. 2017;143(20):66–78.
Crossref Google Scholar
28

He HS, Hao Z, Mladenoff DJ, Shao G, Hu Y, Chang Y. Chang, simulating forest ecosystem response to climate warming incorporating spatial effects in North-Eastern China. J Biogeogr. 2005;32(12):2043–2056.
Crossref Google Scholar
29

Janiesch C, Zschech P, Heinrich KJEM. Machine learning and deep learning. Biomed Sig Process. 2021;31(3):685–695.
Crossref Google Scholar
30
He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. Paper presented at: Proceedings of the IEEE conference on computer vision and pattern recognition; 2016 Jun 23–30; Las Vegas, NV.
Crossref
31

Agarwal M, Gupta SK, Biswas K. Development of a compressed FCN architecture for semantic segmentation using particle swarm optimization. Neural Comput Appl. 2023;35(16):11833–11846.
Crossref Google Scholar
32

Gong H, Liu T, Luo T, Guo J, Feng R, Li J, Ma X, Mu Y, Hu T, Sun YJA, et al. Based on FCN and DenseNet framework for the research of rice pest identification methods. Agronomy. 2023;13(2):410.
Crossref Google Scholar
33

Loebel E, Scheinert M, Horwath M, Humbert A, Sohn J, Heidler K, Liebezeit C. Calving front monitoring at sub-seasonal resolution: A deep learning application to Greenland glaciers. Cryosphere Discuss. 2023;2023:1–21.
Crossref Google Scholar
34
Strudel R, Garcia R, Laptev I, Schmid C. Segmenter: Transformer for semantic segmentation. Paper presented at: Proceedings of the IEEE/CVF International Conference on Computer Vision; 2021 Jun 19–25; online.
Crossref
35

Oršić M, Šegvić S. Efficient semantic segmentation with pyramidal fusion. Pattern Recogn. 2021;110:107611.
Crossref Google Scholar
36

Kang J, Liu L, Zhang F, Shen C, Wang N, Shao LJC. Semantic segmentation model of cotton roots in-situ image based on attention mechanism. Comput Electron Agric. 2021;189: Article 106370.
Crossref Google Scholar
37

Li R, Zheng S, Zhang C, Duan C, Su J, Wang L. Multiattention network for semantic segmentation of fine-resolution remote sensing images. IEEE Trans Geosci Remote Sens. 2021;60(99):1–13.
Crossref Google Scholar
38

Su Y, Cheng J, Bai H, Liu H, He CJRS. Semantic segmentation of very-high-resolution remote sensing images via deep multi-feature learning. Remote Sens. 2022;14(3):533.
Crossref Google Scholar
39

Kim S, Park S, Kwak J, Choi IJEI. Semantic segmentation of seagrass habitat from drone imagery based on deep learning: A comparative study. Eco Inform. 2021;66:101430.
Crossref Google Scholar
40

Zhu D, Qian C, Qu C, He M, Zhang S, Tu Q. An improved SegNet network model for accurate detection and segmentation of car body welding slags. Int J Adv Manuf Technol. 2022;120(1-2):1095–1105.
Crossref Google Scholar
41
Hu X, Naiel MA, Wong A, Lamm M. RUNet: A robust UNet architecture for image super-resolution, Paper presented at: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW); 2019 Jun 16–17; Long Beach, CA.
Crossref
42

Liu F, Wang LJC, Materials B. UNet-based model for crack detection integrating visual explanations. Constr Build Mater. 2022;322:126265.
Crossref Google Scholar
43
Zhang Z. Generalized cross entropy loss for training deep neural networks with noisy labels, Paper presented at: Proceedings of the 32nd International Conference on Neural Information Processing Systems, NIPS’18; 2018 Dec 10–12; Montreal, Canada.
44
Jadon S, A survey of loss functions for semantic segmentation. Paper presented at: IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB);2020 Oct 27–29; Via del Mar, Chile.
Crossref
45
Denk M, Rother K, Zinßer M, Petroll C. Nodal cosine sine material interpolation in multi objective topology optimization with the global criteria method for linear elasto static, heat transfer, potential flow and binary cross entropy sharpening. Paper presented at: Proceedings of the Design Society; 2021 Oct 12–14; Beijing, China.
Crossref
46
Qi Q, Luo Y, Xu Z, Ji S, Yang T. Stochastic optimization of areas under precision-recall curves with provable convergence. ArXiv. 2021. https://doi.org/10.48550/arXiv.2104.08736
47
Yacouby R, Axman D. Probabilistic extension of precision, recall, and F1 score for more thorough evaluation of classification models. Paper presented at: Proceedings of the First Workshop on Evaluation and Comparison of NLP Systems; 2020 Nov 13–16; online.
Crossref
48
Yurtkulu SC, Şahin YH, Unal G. Semantic segmentation with extended DeepLabv3 architecture. Paper presented at: 2019 27th Signal Processing and Communications Applications Conference (SIU). 2019 Apr 24–26; Sivas, Turkey.
Crossref
49
Li X, Sun X, Meng Y, Liang J, Wu F. Dice loss for data-imbalanced NLP tasks. ArXiv. 2019. https://doi.org/10.48550/arXiv.1911.02855
Crossref
Plant Phenomics
Volume 5,
2023
Article number: 0123
DOI: 10.34133/plantphenomics.0123
Cite this article:
Zhang P, Sun X, Zhang D, et al. Lightweight Deep Learning Models for High-Precision Rice Seedling Segmentation from UAV-Based Multispectral Images. Plant Phenomics, 2023, 5: 0123. https://doi.org/10.34133/plantphenomics.0123

144

Views

15

Crossref

12

Web of Science

13

Scopus

0

CSCD

Altmetrics
Received: 15 July 2023
Accepted: 12 November 2023
Published: 30 November 2023
© 2023 Panli Zhang et al. Exclusive licensee Nanjing Agricultural University. No claim to original U.S. Government Works.

Distributed under a Creative Commons Attribution License 4.0 (CC BY 4.0).
Return