Identification of flowering rate of Litchi canopy based on UAV multispectral remote sensing images
Abstract
Abstract: The yield of litchi is closely related to the on-demand and precise operation of litchi trees. For litchi trees with different flowering rates, the quantity of fertilizer application may also vary from tree to tree. In order to reduce labor requirements and improve the efficiency of observing the flowering rate of the litchi canopy, this study combines multispectral remote sensing images with deep learning technology to achieve flowering rate recognition and modeling of the litchi tree canopy in large-scale orchards. This research proposes a technique for merging visible images with multispectral images. The five vegetation index images calculated from remote sensing images of five different wavelength bands were combined with visible three-channel images to derive the most favorable combination of vegetation index channels for identifying the flowering rate of the litchi canopy; the obtained multi-channel images were used as inputs for training, and the Vision Transformer neural network was used to construct a litchi canopy flowering rate recognition model with a normalization method to further improve the accuracy of the model. After normalization, the best results of litchi canopy flowering rate recognition were obtained when RGB was fused with OSAVI and NDVI vegetation indices. Compared with other models, the recognition model constructed based on Vision Transformer achieved an accuracy of 97.22%. This study can accurately identify the flowering rate of the litchi canopy under multispectral remote sensing images and direct the appropriate fertilizing or other production activities, which is helpful to realize the intelligent management of orchards.
Keywords: Multispectral remote sensing image; channel fusion; flowering rate recognition; deep learning
DOI: 10.33440/j.ijpaa.20220501.189
Citation: Chen Y T, Xu J Q, Zhang W F, Qiu S, Lan Y B, Deng X L. Identification of flowering rate of Litchi canopy based on UAV multispectral remote sensing images. Int J Precis Agric Aviat, 2022; 5(1): 21–28.
Full Text:
PDFReferences
Wang Z. 2021 Guangdong Yangxi Litchi Industry Research Report. Friends of Fruit Farmers, 2022, (02): 61–63. (in Chinese)
Shao Y, Ou Z, Huang J, et al. Guangdong lychee industry welcomes the iteration and upgrading. Nanfang Daily, 2022-01-04 (A04). doi: 10.28597/n.cnki.nnfrb.2022.000112. (in Chinese)
Xiong J, Zou X, Liu N, et al. Quality detection technology of litchi fruit picking based on machine vision. Journal of Agricultural Machinery, 2014, 45(07): 54–60. (in Chinese)
Jiang R, Wang C, Shen L, et al. Research on the extraction method of litchi forest canopy information based on high-resolution remote sensing images. Journal of Agricultural Machinery, 2016, 47(09): 17–22. (in Chinese)
Wang J, Chen Y, Zeng Z, et al. Extraction of litchi skin defects based on fully convolutional neural network. Journal of South China Agricultural University, 2018, 39(06): 104–110. (in Chinese)
Xiong J, Liu B, Zhong Z, et al. Segmentation and recognition of litchi mosaics and leaves based on deep semantic segmentation network. Journal of Agricultural Machinery, 2021, 52(06): 252–258. (in Chinese)
Ye J, Qiu W, Yang J, et al. A method for identification of lychee pests based on deep learning. Laboratory Research and Exploration, 2021, 40(06): 29–32. doi: 10.19927/j.cnki.syyt.2021.06.007. (in Chinese)
Lin J, Li J, Yang Z, et al. Estimating litchi flower number using a multicolumn convolutional neural network based on a density map. Precision Agric 23, 1226–1247. doi: 10.1007/s11119-022-09882-7 (2022)
Xie Z, Chen S, Gao G, et al. Evaluation of rapeseed flowering dynamics for different genotypes with UAV platform and machine learning algorithm. Precision Agric. doi: 10.1007/s11119-022-09904-4 (2022).
Karakoç, A., Karabulut, M. Monitoring of Canopy Reflectance Change Based on Flowering Rate. J Indian Soc Remote Sens 48, 1159–1168. (2020). doi: 10.1007/s12524-020-01142-3.
Li H, Lee W S, Wang K, et al. ‘Extended spectral angle mapping(ESAM)’ for citrus greening disease detection using airborne hyperspectral imaging[J]. Precision Agriculture, 2014, 2(15): 162–183. doi: 10.1007/s11119-013-9325-6.
Chen C. Multispectral UAV remote sensing application in agriculture. Modern Agriculture, 2021(10): 61–62. (in Chinese)
Xiao C, Zheng L, Sun H, et al. Estimation of Apple Tree Blooms Based on Aerial Multispectral Images. Chinese Society of Agricultural Engineering. Proceedings of the 2013 Annual Conference of Chinese Agricultural Engineering Society, 2013: 1–4. (in Chinese)
Nogueira M R, de Carvalho Pinto F A, Marçal de Queiroz D, et al. A Novel Vegetation Index for Coffee Ripeness Monitoring Using Aerial Imagery. Remote Sensing, 2021, 13(2): 263. doi: 10.3390/rs13020263.
Deng X, Zeng G, Zhu Z, et al. Classification and feature band extraction of citrus diseased plants based on UAV hyperspectral remote sensing. Journal of South China Agricultural University, 2020, 41(6): 100–108. (in Chinese)
Deng X, Zhu Z, Yang J, et al. Detection of Citrus Huanglongbing Based on Multi-Input Neural Network Model of UAV Hyperspectral Remote Sensing. Remote Sens, 2020 12(17): 2678. doi: 10.3390/rs12172678.
Deng X, Huang Z, Zheng Z, et al. Field detection and classification of citrus Huanglongbing based on hyperspectral reflectance. Computers and Electronics in Agriculture, 2019 167(12): 105006. doi: 10.1016/ j.compag.2019.105006.
Lan Y, Zhu Z, Deng X, et al. Monitoring and classification of citrus Huanglongbing based on UAV hyperspectral remote sensing. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2019, 35(3): 92–100. (in Chinese)
Lan Y, Huang Z, Deng X, et al. Comparison of machine learning methods for citrus greening detection on UAV multispectral images. Computers and Electronics in Agriculture, 2020, 171(5): 105234. doi: 10.1016/ j.compag.2020.105234.
Ma L, Huang C. Application of UAV multi-spectral technology in smart agriculture. Modern manufacturing technology and equipment, 2021, 57(11): 163–165. doi: 10.16107/j.cnki.mmte.2021.0888. (in Chinese)
Yang S, Zheng Y, Liu X, et al. Discrimination of Grape Ripeness of Snake Dragon Ball Based on Near-ground Multispectral Image of UAV. Spectroscopy and Spectral Analysis, 2021, 41(10): 3220–3226. (in Chinese)
Zhang X. Convolutional neural network spectral analysis method and its application in agricultural product quality detection. Zhejiang University, 2021. doi: 10.27461/d.cnki.gzjdx.2021.001151. (in Chinese)
Wu G, Peng Y, Zhou G, et al. Research on the identification method of corn crop nutritional status based on multispectral imaging and convolutional neural networ. Smart Agriculture (Chinese and English), 2020, 2(01): 111–120. (in Chinese)
Fabiano F, Carlos H, Francisco G, et al. Detection of Drechslera avenae (Eidam) Sharif [Helminthosporium avenae (Eidam)] in Black Oat Seeds (Avena strigosa Schreb) Using Multispectral Imaging. Sensors, 2020, 20(12). doi: 10.3390/s20123343.
Fawakherji M, Potena C, Pretto A, et al. Multi-spectral image synthesis for crop/weed segmentation in precision farming. Robotics and Autonomous Systems, 2021(prepublish). doi: 10.1016/j.robot.2021.103861.
Anatoly A G, Yoram J K, Mark N M. Use of a green channel in remote sensing of global vegetation from EOS-MODIS. Remote Sensing of Environment, 1996, 58(3): 289–298. doi: 10.1016/S0034-4257(96)00072-7.
Thenkabail P S, Smith R B , Pauw E D. Hyperspectral vegetation indices for determining agricultural crop characteristics. Remote Sens. Environ, 2000, 71(2): 158–182
Fitzgerald G J , Rodriguez D, Christensen L K, et al. Spectral and thermal sensing for nitrogen and water status in rainfed and irrigated wheat environments. Precis. Agric., 7 (2006), pp. 233–248. doi: 10.1007/ s11119-006-9011-z.
Haboudane D, Miller J R, Pattey E, et al. Strachan. Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop canopies: Modeling and validation in the context of precision agriculture. Remote Sens. Environ., 90 (2004), pp. 337–352. doi: 10.1016/ j.rse.2003.12.013.
Rondeaux G, Steven M, Baret F. Optimization of soil-adjusted vegetation indices. Remote Sens. Environ., 55 (1996), pp. 95–107. doi: 10.1016/ 0034-4257(95)00186-7.
Mo J, Lan Y, Yang D, et al. Deep Learning-Based Instance Segmentation Method of Litchi Canopy from UAV-Acquired Images. REMOTE SENSING, 2021, 13(19).
Alexey D, Lucas B, Alexander K, et al. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. Computer Vision and Pattern Recognition, 2021. doi: 10.48550/arXiv.2010.11929.
Refbacks
- There are currently no refbacks.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.