Remote Sensing and GIS Applications in Environmental Sciences

Strawberry Crop Classification Using Sentinel-2 Satellite Imagery and a CNN Deep Learning Algorithm: A Phenology-Based Local Approach

Document Type : Original Article

Authors

1 Imam Hossein Comprehensive University

2 Jihad University of Kurdistan

3 Shahid Beheshti University

4 Hekmat Higher Educational Institution

10.22034/rsgi.2025.67701.1140

Abstract

The aim of this study is to develop and evaluate a hybrid deep learning-based model for accurate classification of strawberry crops using Sentinel-2 satellite time-series imagery and indigenous phenological data in Kurdistan Province. The primary motivation for designing the model lies in the spatial identification of crops with limited distribution and specific growth cycles.
Ground truth data were collected and labeled during the 2022 growing season using GPS and the regional agricultural calendar. Selected Sentinel-2 images temporally aligned with the phenological stages of strawberry growth underwent geometric, radiometric, and atmospheric preprocessing. These data were then fed into a Convolutional Neural Network (CNN) consisting of three convolutional layers, one pooling layer, two fully connected layers, and a Softmax output function. To address class imbalance, class weighting techniques and grid search optimization were employed. Data augmentation was also utilized to enhance the model's generalizability. The training process was conducted using the Adam optimizer with a learning rate of 0.0001 over 150 epochs. The proposed model successfully classified 15 land use/land cover classes with an overall accuracy of 96.57% and a Kappa coefficient of 0.8582. For the target class—strawberry—the F1-score reached 86.4%, indicating a favorable balance between precision and recall. The model demonstrated strong performance in identifying crops with limited cultivated areas and irregular spatial structures, although accuracy declined in classes such as greenhouses and rangelands due to high spectral heterogeneity. Error analysis revealed that most misclassifications occurred at spectrally similar boundaries and during climatic fluctuations in the growth period.

Keywords

Main Subjects

هدف: هدف این پژوهش، توسعه و ارزیابی یک مدل ترکیبی مبتنی بر یادگیری عمیق جهت طبقه‌بندی دقیق محصول توت‌فرنگی با استفاده از تصاویر سری‌زمانی ماهواره سنتینل-۲ و داده‌های فنولوژیکی بومی در استان کردستان است. تمرکز بر شناسایی مکانی محصولات با پراکنش محدود و چرخه رشد خاص، انگیزه اصلی طراحی مدل در این تحقیق بوده است. علاوه بر این، تلاش شده است تا کارایی مدل در شرایط مکانی متنوع و اقلیمی منطقه‌ای نیز مورد بررسی قرار گیرد.

روش پژوهش: ابتدا داده‌های زمینی در فصل زراعی ۱۴۰۱ از طریق GPS و تقویم زراعی منطقه گردآوری و برچسب‌گذاری شدند. سپس تصاویر منتخب سنتینل-۲ که از لحاظ زمانی با مراحل فنولوژیکی رشد توت‌فرنگی هم‌پوشانی داشتند، تحت پیش‌پردازش‌های هندسی، رادیومتریکی و اتمسفری قرار گرفتند. این داده‌ها به‌عنوان ورودی به یک شبکه عصبی کانولوشنی با سه لایه کانولوشن، یک لایه Pooling، دو لایه تماماً متصل و تابع خروجی Softmax معرفی شدند. برای مقابله با نامتوازنی کلاس‌ها، از تکنیک وزن‌دهی کلاس و بهینه‌سازی پارامترها با روش جستجوی شبکه‌ای استفاده شد. همچنین از داده‌افزایی برای بهبود تعمیم‌پذیری مدل بهره گرفته شد. فرایند آموزش با الگوریتم Adam و نرخ یادگیری 0001/0 در طی 150 اپک انجام شد.

Ahangarha, M., Saadat Seresht, M., Shahhoseini, R., & Seyyedi, S. (2020). Crop land change monitoring based on deep learning algorithm using multi-temporal hyperspectral images. Journal of Geomatics Science and Technology, 10(2), 79–89. (In Persian) https://dor.isc.ac/dor/20.1001.1.2322102.1399.10.2.6.9
Alipour, F., Aghkhani, M., Abbaspour, F. M., & Sepehr, A. (2015). Demarcation and estimation of agricultural lands using ETM+ imagery data (case study: Astan Ghods Razavi great farm). (In Persian). https://dor.isc.ac/dor/20.1001.1.22286829.1393.4.2.12.8
An, Q., Wang, K., Li, Z., Song, C., Tang, X., & Song, J. (2022). Real-time monitoring method of strawberry fruit growth state based on YOLO improved model. IEEE Access, 10, 124363–124372. https://doi.org/10.1109/ACCESS.2022.3220234
Anjom, F. K., Vougioukas, S., & Slaughter, D. C. (2018). Development and application of a strawberry yield-monitoring picking cart. Computers and Electronics in Agriculture, 155(2), 400–411. http://dx.doi.org/10.1016/j.compag.2018.10.038
Bargiel, D. (2017). A new method for crop classification combining time series of radar images and crop phenology information. Remote Sensing of Environment, 198, 369–383. https://doi.org/10.1016/j.rse.­2017.06.022
Chang, C.-I. (2007). Hyperspectral data exploitation: Theory and applications. John Wiley & Sons. DOI:10.1002/0470124628
Chen, Y., Wu, X., & Xu, Q. (2024). A hybrid deep learning model based on Bi-LSTM for crop type classification in smallholder farms using multi-temporal satellite data. PLoS ONE, 19(3), e0299350. https://doi.org/10.1371/journal.pone.0299350
Constante, P. N., Gordon, A., Chang, O., & Pruna, E. (2016). Artificial vision techniques to optimize strawberry's industrial classification. IEEE Latin America Transactions, 14(6), 2576–2581. http://dx.doi.org/10.1109/TLA.2016.7555221
Csillik, O., & Belgiu, M. (2017). Cropland mapping from Sentinel-2 time series data using object-based image analysis. In Proceedings of the 20th AGILE International Conference on Geographic Information Science Societal Geo-Innovation Celebrating (Vol. 9, p. 20). Wageningen, The Netherlands.
Du, X., Cheng, H., Ma, Z., Lu, W., Wang, M., Meng, Z., Jiang, C., & Hong, F. (2023). DSW-YOLO: A detection method for ground-planted strawberry fruits under different occlusion levels. Computers and Electronics in Agriculture, 214, 108304. https://doi.org/10.1016/j.compag.2023.108304
Gadiraju, K. K., & Vatsavai, R. R. (2020). Comparative analysis of deep transfer learning performance on crop classification. In Proceedings of the 9th ACM SIGSPATIAL International Workshop on Analytics for Big Geospatial Data (pp. 1–8). https://doi.org/10.1145/3423336.3431369
Heidari,M. , tashakori,M. and Valipour,A. (2016). Investigating the local utilization of forest in north Zagros (Case study: Baneh Region). Iranian Journal of Forest, 8(3), 313-331. (In Persian) https://www.ijf-isaforestry.ir/article_42271.html?lang=en
Kang, Y., Khan, S., & Ma, X. (2009). Climate change impacts on crop yield, crop water productivity and food security: A review. Progress in Natural Science, 19(12), 1665–1674. https://doi.org/10.1016/j.pnsc.­2009.­08.001
Khatami, R., Mountrakis, G., & Stehman, S. V. (2017). Mapping per-pixel predicted accuracy of classified remote sensing images. Remote Sensing of Environment, 191, 156–167. http://dx.doi.org/10.1016/j.rse.2017.­01.025
Kingma, D. P., & Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980. https://doi.org/10.48550/arXiv.1412.6980
Li, Z., Wang, J., Gao, G., Lei, Y., Zhao, C., Wang, Y., Bai, H., Liu, Y., Guo, X., & Li, Q. (2024). SGSNet: A lightweight deep learning model for fast and accurate detection of strawberry growth stages. Frontiers in Plant Science, 15, 1491706. https://doi.org/10.3389/fpls.2024.1491706
Lyu, H., Lu, H., & Mou, L. (2016). Learning a transferable change rule from a recurrent neural network for land cover change detection. Remote Sensing, 8(6), 506. https://doi.org/10.3390/rs8060506
Saleem, N., Balu, A., Jubery, T. Z., Singh, A., Singh, S. K., Sarkar, S., & Ganapathysubramanian, B. (2024). Class-specific data augmentation for plant stress classification in soybean. bioRxiv. https://doi.org/10.1101/2024.06.18.549378
 
Wang, C., Wang, H., Han, Q., Zhang, Z., Kong, D., & Zou, X. (2024). Strawberry Detection and Ripeness Classification Using YOLOv8+ Model and Image Processing Method. Agriculture, 14(5), 751. https://doi.org/10.3390/agriculture14050751
Xu, J., et al. (2020). DeepCropMapping: A multi-temporal deep learning approach with improved spatial generalizability for dynamic corn and soybean mapping. Remote Sensing of Environment, 247, 111946. https://doi.org/10.1016/j.rse.2020.111946
Zheng, B., Myint, S. W., Thenkabail, P. S., & Aggarwal, R. M. (2015). A support vector machine to identify irrigated crop types using time-series Landsat NDVI data. International Journal of Applied Earth Observation and Geoinformation, 34, 103–112. https://doi.org/10.1016/j.jag.2014.07.002
Remote Sensing and GIS Applications in Environmental Sciences
Volume 6, Issue 18
April 2026
Pages 96-78
  • Receive Date: 05 June 2025
  • Revise Date: 18 October 2025
  • Accept Date: 30 December 2025