Mapping rice-crayfish fields using high-resolution imagery and CNN-Transformer network: a case study of the hinterland of Jianghan Plain, China

Abstract

The rice-crayfish coculture system involves growing rice and crayfish together in dedicated fields, providing important economic and ecological benefits. Accurate and timely instance-level monitoring of rice-crayfish fields (RCFs) is essential for tracking rice growth, estimating yield, and managing water resources. Previous deep learning-based approaches, relying on semantic segmentation and extra post-processing for RCF instance isolation, which can introduce propagate errors, leading to boundary inaccuracies and reduced reliability in real-world applications. To address these issues, we propose EMC-YOLO, an end-to-end instance segmentation framework that integrates an efficient vision transformer with a multi-scale, contour-guided YOLO architecture. By unifying global context modeling, multi-scale representation, and boundary refinement within a single pipeline, the proposed method performs direct instance-aware mask prediction. This design eliminates error-prone post-processing and enables precise delineation of individual RCFs across diverse scales, shapes, and crop stages. Experimental results highlight that the proposed method achieves a box mAP (mean average precision) of 0.954 and a mask mAP of 0.952, surpassing YOLOv8n-seg by 1.6 and 2.5 percentage points, respectively. This robust performance enables reliable extraction of RCF instances under variable field conditions, establishing EMC-YOLO as a powerful solution for high-precision, fine-scale mapping of rice-crayfish coculture systems and providing actionable technical support for smart agriculture implementation.

Graphical abstract

Keywords

rice-crayfish field mapping / high-resolution imagery / instance segmentation / multi-scale feature representation / contour enhancement

References

[1] Bolya D, Zhou C, Xiao F, Lee Y J (2019). YOLACT: real-time instance segmentation. In: Proceedings of 2019 IEEE/CVF International Conference on Computer Vision (ICCV). IEEE, 9156–9165

[2] Buslaev A, Iglovikov V I, Khvedchenya E, Parinov A, Druzhinin M, Kalinin A A (2020). Albumentations: fast and flexible image augmentations.Information (Basel), 11(2): 125

[3] Cai Z, Vasconcelos N (2021). Cascade R-CNN: high quality object detection and instance segmentation.IEEE Trans Pattern Anal Mach Intell, 43(5): 1483–1498

[4] Cai Z, Wei H, Hu Q, Zhou W, Zhang X, Jin W, Wang L, Yu S, Wang Z, Xu B, Shi Z (2023). Learning spectral-spatial representations from VHR images for fine-scale crop type mapping: a case study of rice-crayfish field extraction in South China.ISPRS J Photogramm Remote Sens, 199: 28–39

[5] Chen K, Pang J, Wang J, Xiong Y, Li X, Sun S, Feng W, Liu Z, Shi J, Quyang W, Loy C C, Lin D (2019). Hybrid task cascade for instance segmentation. In: Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 4969–4978

[6] Chen W T, Vong Y J, Kuo S Y, Ma S, Wang J (2024). RobustSAM: segment anything robustly on degraded images. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 4081–4091

[7] Chen Y, Yu P, Chen Y, Chen Z (2022). Spatiotemporal dynamics of rice–crayfish field in Mid-China and its socioeconomic benefits on rural revitalisation.Appl Geogr, 139: 102636

[8] Chen Y, Yuan X, Wu R, Wang J, Hou Q, Cheng M M (2023). YOLO-MS: rethinking multi-scale representation learning for real-time object detection. arXiv:2308.05480

[9] Ding L, Goshtasby A (2001). On the Canny edge detector.Pattern Recognit, 34(3): 721–725

[10] Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner, T, Dehghani M, Minderer M, Heigold G, Gelly S, Uszkoreit J, Houlsby, N (202202). An image is worth 16 × 16 words: transformers for image recognition at scale. In: Proceedings of the International Conference on Learning Representations. ICLR: 611–632

[11] Gao S, Cheng M, Zhao K, Zhang X, Yang M, Torr P (2021). Res2Net: a new multi-scale backbone architecture.IEEE Trans Pattern Anal Mach Intell, 43(2): 652–662

[12] Guo M, Xiong F, Zhao B, Huang Y, Xie Z, Wu L, Chen X, Zhang J (2024a). TDEGAN: a texture-detail-enhanced dense generative adversarial network for remote sensing image super-resolution.Remote Sens (Basel), 16(13): 2312

[13] Guo M, Zhang H, Huang Y, Xie Z, Wu L, Zhang J (2024b). Shadow removal method for high-resolution aerial remote sensing images based on region group matching.Expert Syst Appl, 255: 124739

[14] Han D, Pan X, Han Y, Song S, Huang G (20232023). FLatten Transformer: vision transformer using focused linear attention. In: Proceedings of 2023 IEEE/CVF International Conference on Computer Vision (ICCV). IEEE, 5938–5948

[15] Han J, Zhang Z, Luo Y, Cao J, Zhang L, Zhuang H, Cheng F, Zhang J, Tao F (2022). Annual paddy rice planting area and cropping intensity datasets and their dynamics in the Asian monsoon region from 2000 to 2020.Agric Syst, 200: 103437

[16] Han Y, Huang J, Ling F, Qiu J, Liu Z, Li X, Chang C, Chi H (2023a). Dynamic mapping of inland freshwater aquaculture areas in Jianghan Plain, China.IEEE J Sel Top Appl Earth Obs Remote Sens, 16: 4349–4361

[17] He K, Gkioxari G, Dollár P, Girshick R (2017). Mask R-CNN. In: Proceedings of 2017 IEEE International Conference on Computer Vision (ICCV). IEEE, 2980–2988

[18] Hossain M D, Chen D (2019). Segmentation for object-based image analysis (OBIA): a review of algorithms and challenges from remote sensing perspective.ISPRS J Photogramm Remote Sens, 150: 115–134

[19] Howard A, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, Andreetto M, Adam H (2017). MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. arXiv:1704.04861

[20] Hu J, Shen L, Sun G (2018). Squeeze-and-excitation networks. In: Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, 7132–7141

[21] Huang Z, Huang L, Gong Y, Huang C, Wang X (2019). Mask Scoring R-CNN. In: Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 6402–6411

[22] Jiang Y, Cao C (2021). Crayfish–rice integrated system of production: an agriculture success story in China. A review.Agron Sustain Dev, 41(5): 68

[23] Jocher G (2022). YOLOv5 by Ultralytics (version 5.7.0). Retrieved March 24, 2025

[24] Jocher G, Chaurasia A, Qiu J (2023). YOLOv8 by Ultralytics (version 8.0.0). Retrieved March 24, 2025

[25] Jocher G, Chaurasia A, Qiu J (2024). YOLO11 by Ultralytics (version 8.3.0). Retrieved March 24, 2025

[26] Joshi A, Pradhan B, Gite S, Chakraborty S (2023). Remote-sensing data and deep-learning techniques in crop mapping and yield prediction: a systematic review.Remote Sens (Basel), 15(8): 2014

[27] Jung A B, Wada K, Crall J, Tanaka S, Graving J, Yadav S, Banerjee J, Vecsei G, Kraftt A, Rui Z, Borovec J, Vallentin C (2020). imgaug (version 0.4.0). Retrieved February 27, 2025

[28] Kang M, Ting C M, Ting F, Phan R (2024). ASF-YOLO: a novel YOLO model with attentional scale sequence fusion for cell instance segmentation.Image Vis Comput, 147: 105057

[29] Kirillov A, Mintun E, Ravi N, Mao H, Rolland C, Gustafson L, Xiao T, Whitehead S, Berg A, Lo W, Dollár P, Girshick R (2023). Segment anything. In: Proceedings of 2023 IEEE/CVF international conference on computer vision (ICCV). IEEE, 3992–4003

[30] Kirillov A, Wu Y, He K, Girshick R (2020). PointRend: image segmentation as rendering. In: Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 9796–9805

[31] Li Y, Ku X, Wang S, Pan G, Xu Q, Liu H, Li J (2024). Evaluation of annual production and economic benefits of tropical rice-crayfish coculture system in China.Agric Syst, 217: 103936

[32] Liu S, Qi L, Qin H, Shi J, Jia J (2018). Path aggregation network for instance segmentation. In: Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, 8759–8768

[33] Liu X, Peng H, Zheng N, Yang Y, Hu H, Yuan Y (2023). EfficientViT: memory efficient vision transformer with cascaded group attention. In: Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 14420–14430

[34] Liu Y, Zhao W, Chen S, Ye T (2021). Mapping crop rotation by using deeply synergistic optical and SAR time series.Remote Sens (Basel), 13(20): 4160

[35] Luo F, Zhang Y, Zhao X (2024a). Understanding the spatiotemporal dynamics and influencing factors of the rice-crayfish field in Jianghan Plain, China.Remote Sens (Basel), 16(9): 1541

[36] Luo M, Zhang T, Wei S, Ji S (2024b). SAM-RSIS: progressively adapting SAM with box prompting to remote sensing image instance segmentation.IEEE Trans Geosci Remote Sens, 62: 1–14

[37] Ma S, Wang D, Yang H, Hou H, Li C, Li Z (2023). A bi-temporal-feature-difference- and object-based method for mapping rice-crayfish fields in Sihong, China.Remote Sens (Basel), 15(3): 658

[38] Meituan (2023). YOLOv6-Segmentation (version 0.4.1). Retrieved March 24, 2025

[39] Munawar R (2022). YOLOv7-Segmentation. Retrieved February 27, 2025

[40] National Fisheries Technology Extension Center (NFTEC) (2017). Report on the Development of China’s Crawfish Industry (2017).China Fisheries, (7): 8–17

[41] National Fisheries Technology Extension Center (NFTEC), China Society of Fisheries (CSF) (2024). Report on the Development of China’s Crawfish Industry (2024).China Fisheries, 584(7): 14–20

[42] Niu K, Wang C, Xu J, Yang C, Zhou X, Yang X (2023). An Improved YOLOv5s-Seg Detection and Segmentation Model for the Accurate Identification of Forest Fires Based on UAV Infrared Image.Remote Sens (Basel), 15(19): 4694

[43] Ronneberger O, Fischer P, Brox T (2015). U-Net: convolutional networks for biomedical image segmentation. In: Proceedings of Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. Cham: Springer International Publishing, 234–241

[44] Sandler M, Howard A, Zhu M, Zhmoginov A, Chen L (2018). MobileNetV2: Inverted Residuals and Linear Bottlenecks. In: Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, 4510–4520

[45] Shen A, Zhu Y, Angelov P, Jiang R (2024). Marine debris detection in satellite surveillance using attention mechanisms.IEEE J Sel Top Appl Earth Obs Remote Sens, 17: 4320–4330

[46] Tian Z, Shen C, Wang X, Chen H (2021). BoxInst: high-performance instance segmentation with box annotations. In: Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 5439–5448

[47] Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez A N, Kaiser L Polosukhin l, Guyon l, Luxburg U V, Bengio S, Wallach H, Fergus R, Vishwanathan S, Garnett R (2017). Attention is all you need. In: Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook, NY: Curran Associates Inc., 6000–6010

[48] Wang C Y, Bochkovskiy A, Liao H Y M (2023). YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors. In: Proceedings of 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 7464–7475

[49] Wang C Y, Yeh I H, Mark Liao H Y (2024). YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information. In: Proceedings of Computer Vision – ECCV 2024: 18th European Conference, Milan, Italy, September 29–October 4, 2024, Proceedings, Part XXXI. Berlin, Heidelberg: Springer-Verlag, 1–21

[50] Wang X, Zhang R, Kong T, Li L, Shen C (2020). SOLOv2: dynamic and fast instance segmentation. In: Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook, NY, USA: Curran Associates Inc

[51] Wang X, Zhang R, Shen C, Kong T, Li L (2022). SOLO: a simple framework for instance segmentation.IEEE Trans Pattern Anal Mach Intell, 44(11): 8587–8601

[52] Wei H, Hu Q, Cai Z, Yang J, Song Q, Yin G, Xu B (2021b). An object- and topology-based analysis (OTBA) method for mapping rice-crayfish fields in South China.Remote Sens (Basel), 13(22): 4666

[53] Wei Y, Lu M, Yu Q, Xie A, Hu Q, Wu W (2021a). Understanding the dynamics of integrated rice–crawfish farming in Qianjiang county, China using Landsat time series images.Agric Syst, 191: 103167

[54] Weiss M, Jacob F, Duveiller G (2020). Remote sensing for agricultural applications: a meta-review.Remote Sens Environ, 236: 111402

[55] Woo S, Park J, Lee J Y, Kweon I S (2018). CBAM: convolutional block attention module. In: Proceedings of Computer Vision – ECCV 2018. Cham: Springer International Publishing, 3–19

[56] Xia T, He Z, Cai Z, Wang C, Wang W, Wang J, Hu Q, Song Q (2022). Exploring the potential of Chinese GF-6 images for crop mapping in regions with complex agricultural landscapes.Int J Appl Earth Obs Geoinf, 107: 102702

[57] Xia T, Ji W, Li W, Zhang C, Wu W (2021). Phenology-based decision tree classification of rice-crayfish fields from Sentinel-2 imagery in Qianjiang, China.Int J Remote Sens, 42(21): 8124–8144

[58] Xu G, Liao W, Zhang X, Li C, He X, Wu X (2023b). Haar wavelet downsampling: a simple but effective downsampling module for semantic segmentation.Pattern Recognit, 143: 109819

[59] Xu Q, Dai L, Zhou Y, Dou Z, Gao W, Yuan X, Gao W, Zhang H (2023a). Effect of nitrogen application on greenhouse gas emissions and nitrogen uptake by plants in integrated rice-crayfish farming.Sci Total Environ, 905: 167629

[60] Yang H, Wang D, Xin J, Qian H, Li C, Wang Y, Tan Y, Dai J, Zhao H, Li Z (2024a). Integrating multidimensional feature indices and phenological windows for mapping cropping patterns in complex agricultural landscape regions.IEEE J Sel Top Appl Earth Obs Remote Sens, 17: 7448–7459

[61] Yang J, Hu Q, Li W, Song Q, Cai Z, Zhang X, Wei H, Wu W (2024b). An automated sample generation method by integrating phenology domain optical-SAR features in rice cropping pattern mapping.Remote Sens Environ, 314: 114387

[62] Yasir M, Liu S, Pirasteh S, Xu M, Sheng H, Wan J, de Figueiredo A P, Aguilar F J, Li J (2024a). YOLOShipTracker: tracking ships in SAR images using lightweight YOLOv8.Int J Appl Earth Obs Geoinf, 134: 104137

[63] Yasir M, Liu S, Xu M, Wan J, Shah N, Qamar U I, Kinh B D (2024b). SwinYOLOv7: robust ship detection in complex synthetic aperture radar images.Appl Soft Comput, 160: 111704

[64] Yuan P, Wang J, Guo C, Guo Z, Guo Y, Cao C (2022). Sustainability of the rice-crayfish farming model in waterlogged land: a case study in Qianjiang County, Hubei Province, China.J Integr Agric, 21(4): 1203–1214

[65] Yuan Q, Zou S, Wang H, Luo W, Zheng X, Liu L, Meng Z (2024). A lightweight pine wilt disease detection method based on vision transformer-enhanced YOLO.Forests, 15(6): 1050

[66] Zambre Y V, Rajkitkul E, Mohan A, Peeples J (2024) Spatial transformer network YOLO model for agricultural object detection. arXiv: 2407.21652

[67] Zha H, Li H, Zhu Y, Wang S, He Y (2023). Extraction of rice and shrimp co-cultivation farming fields using edge-assisted segmentation network.Trans Chin Soc Agric Eng, 39(19): 154–164

[68] Zhao Z, He P (2024). YOLO-U: multi-task model for vehicle detection and road segmentation in UAV aerial imagery.Earth Sci Inform, 17(4): 3253–3269

[69] Zou R, Liu J, Pan H, Tang D, Zhou R (2024). An improved instance segmentation method for fast assessment of damaged buildings based on post-earthquake UAV images.Sensors (Basel), 24(13): 4371

Rights & permissions

Higher Education Press