Identification method of plot-scale rice cultivation in southern hilly regions

doi:10.13733/j.jcam.issn.20955553.2024.12.027

Abstract

Abstract:

Aiming at the hilly area with cloudy rain and complex plot distribution in southern China, a new method for multi-season rice plot level structure information recognition is proposed. By utilizing high-resolution optical imagery and time-series Sentinel-1A SAR imagery, integrated with the Psi-Net model for multi-task semantic segmentation, the study analyzes the relationship between rice growth phenology and backscatter coefficients. The method achieves precise extraction of multi-season rice planting distributions through thresholding, validated in Pucheng County, Fujian Province. Results demonstrate superior performance in shape preservation and boundary accuracy, with Hausdorff distance of 21.368, notably better than the single-task U-Net network. Overall, the method achieves 88.6% and 87.7% accuracies for mid-season and late-season rice identification, respectively, with Kappa coefficients of 0.752 and 0.738. These findings underscore the significant application potential and practical value of the proposed approach for rice planting plot recognition under complex climatic and terrain conditions in southern regions.

Key words: southern hilly areas, rice recognition, SAR image, time series, high resolution, plot

摘要：

针对南方多云雨且地块分布复杂的丘陵地区，提出一种适用该地区的多季水稻地块级结构信息识别方法。利用一期亚米级光学影像与时序Sentinel-1A SAR影像，结合Psi-Net模型进行多任务语义分割，分析水稻生长物候特征与后向散射系数的关系，最终通过阈值法实现对多季水稻种植分布信息的准确提取，并在福建省浦城县进行验证。结果表明，该方法在形状保持能力和识别精度上表现优异。中稻和晚稻种植地块的边界与地面真实值吻合度较高，豪斯多夫距离为21.368，明显优于单一任务的U-Net网络。中稻和晚稻识别的总体精度分别达到88.6%和87.7%，Kappa系数分别为0.752和0.738。所提方法在南方复杂气候和地形条件下的水稻种植地块识别上具有显著应用潜力和实用价值。

关键词: 南方丘陵地区, 水稻识别, SAR影像, 时间序列, 高分辨率, 地块

CLC Number:

TP79

Wu Ruijiao, Chen Guangjian. Identification method of plot-scale rice cultivation in southern hilly regions[J]. Journal of Chinese Agricultural Mechanization, 2024, 45(12): 181-186.

吴瑞姣, 陈光剑. 地块尺度的南方丘陵地区水稻种植识别方法[J]. 中国农机化学报, 2024, 45(12): 181-186.

References

［1］　林志慧, 刘宪锋, 陈瑛, 等. 水—粮食—能源纽带关系研究进展与展望［J］. 地理学报, 2021, 76(7): 1591-1604.
Lin Zhihui, Liu Xianfeng, Chen Ying, et al. Waterfoodenergy nexus: Progress, challenges and prospect ［J］. Acta Geographica Sinica, 2021, 76(7): 1591-1604.
［2］　吴炳方, 张淼, 曾红伟, 等. 全球农情遥感速报系统20年［J］. 遥感学报, 2019, 23(6): 1053-1063.
Wu Bingfang, Zhang Miao, Zeng Hongwei, et al. Twenty years of crop watch: Progress and prospect ［J］. National Remote Sensing Bulletin, 2019, 23(6): 1053-1063.
［3］　谭深, 吴炳方, 张鑫. 基于Google Earth Engine与多源遥感数据的海南水稻分类研究［J］. 地球信息科学学报, 2019, 21(6): 937-947.
Tan Shen, Wu Bingfang, Zhang Xin. Mapping paddy rice in the Hainan Province using both google earth engine and remote sensing images ［J］. Journal of Geoinformation Science, 2019, 21(6): 937-947.
［4］　杨知. 基于极化SAR的水稻物候期监测与参数反演研究［D］. 北京: 中国科学院大学, 2017.
［5］　杨颖频, 吴志峰, 骆剑承, 等. 时空协同的地块尺度作物分布遥感提取［J］. 农业工程学报, 2021, 37(7): 166-174.
Yang Yinpin, Wu Zhifeng, Luo Jiangcheng, et al. Parcelbased crop distribution extraction using the spatiotemporal collaboration of remote sensing data ［J］. Transactions of the Chinese Society of Agricultural Engineering, 2021, 37(7): 166-174.
［6］　李恒凯, 贺明华, 王秀丽. 时序SAR与光学影像的南方丘陵区水稻种植识别［J］. 测绘科学, 2023, 48(5): 78-86.
Li Hengkai, He Minghua, Wang Xiuli. Rice extraction in southern hilly areas by fusing timeseries SAR features and optical images ［J］. Science of Surveying and Mapping, 2023, 48(5): 78-86.
［7］　Kristofer L, Prasad K V, Tuan V T, et al. Mapping double and single crop paddy rice with Sentinel-1A at varying spatial scales and polarizations in Hanoi, Vietnam ［J］. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11(2): 498-512.
［8］　Li H, Fu D, Huang C, et al. An approach to highresolution rice paddy mapping using timeseries Sentinel-1 SAR data in the Mun River Basin, Thailand ［J］. Remote Sensing, 2020, 12(23): 3959.
［9］　查东平, 蔡海生, 张学玲, 等. 基于多时相Sentinel-1水稻种植范围提取［J］. 自然资源遥感, 2022, 34(3): 184-195.
Zha Dongping, Cai Haisheng, Zhang Xueling, et al. Extraction of paddy fields using multitemporal Sentinel-1images ［J］. Remote Sensing for Natural Resources, 2022, 34(3): 184-195.
［10］　张征云, 江文渊, 张彦敏, 等. 基于哨兵SAR数据和多光谱数据的水稻识别研究［J］. 生态与农村环境学报, 2023, 39(4): 556-564.
Zhang Zhengyun, Jiang Wenyuan, Zhang Yanmin, et al. Research on rice area extraction based on Sentinel SAR data and multispectral data ［J］. Journal of Ecology and Rural Environment, 2023, 39(4): 556-564.
［11］　乔树亭, 叶回春, 黄文江, 等. 基于Sentinel-1/2影像的水稻种植面积提取方法研究——以三江平原为例［J］. 遥感技术与应用, 2023, 38(1): 78-89.
Qiao Shuting, Ye Huichun, Huang Wenjiang, et al. Study on extraction method of rice planting area based on Sentinel-1/2 image: A case study of Sanjiang Plain ［J］. Remote Sensing Technology and Application, 2023, 38(1): 78-89.
［12］　刘剑, 王冬至. 基于SAR影像的广东省垦造水田监测应用研究［J］. 测绘通报, 2021(12): 79-82.
Liu Jian, Wang Dongzhi. Study on the application of SAR images in monitoring the build paddy fields in Guangdong province ［J］. Bulletin of Surveying and Mapping, 2021（12）: 79-82.
［13］　左坚, 钟凯文, 许剑辉, 等. 多时相SAR数据下的岭南早稻提取与时空变化分析［J］. 测绘科学, 2023, 48(1): 91-99.
Zuo Jian, Zhong Kaiwen, Xu Jianhui, et al. Analysis on early rice distribution and spatiotemporal changes based on multitemporal SAR data in Lingnan ［J］. Science of Surveying and Mapping, 2023, 48(1): 91-99.
［14］　Cai Y, Lin H, Zhang M. Mapping paddy rice by the objectbased random forest method using time series Sentinel-1/Sentinel-2 data ［J］. Advances in Space Research, 2019, 64(11): 2233-2244.
［15］　曹庆安, 冷鹏. Sentinel-1A时序数据结合多源信息的早稻种植面积监测研究［J］. 江西农业学报, 2023, 35(3): 203-211.
Cao Qingan, Leng Peng. Monitoring of early rice planting area based on Sentinel-1A time series data and multisource information ［J］. Acta Agriculturae Jiangxi, 2023, 35(3): 203-211.
［16］　Murugesan B, Sarveswaran K, Shankaranarayana S M, et al. Psi-Net: Shape and boundary aware joint multitask deep network for medical image segmentation ［C］. 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE, 2019: 7223-7226.
［17］　张晗, 任志远. 基于Whittaker滤波的陕西省植被物候特征［J］. 中国沙漠, 2015, 35(4): 901-906.
Zhang Han, Ren Zhiyuan. Remote sensing analysis of vegetation phenology characteristics in Shaanxi Province based on Whittaker smoother method ［J］. Journal of Desert Research, 2015, 35(4): 901-906.

[1]	Shi Aiping, Qian Zhenwei, Li Yinghao, Feng Liang. Study on prediction method of downy mildew in wine grapes based on GA-LSTM [J]. Journal of Chinese Agricultural Mechanization, 2023, 44(10): 144-151.
[2]	Mao Xiaojuan, Bao Tong, Xun Guanglian, Li Decui, Wang Baojia, Ren Ni. Prediction of temperature in the greenhouse of vegetable growing based on GWO-LSTM [J]. Journal of Chinese Agricultural Mechanization, 2023, 44(1): 116-123.