Construction of a remote sensing image inversion model for saline-alkali land based on deep learning

doi:10.14088/j.cnki.issn0439-8114.2025.06.031

Abstract

Abstract: To address the inversion problem of remote sensing images of saline-alkali land in the Tailan River Irrigation District of Wensu County, Aksu Prefecture, Xinjiang, an image classification model based on the PointRend network integrated with the CoTAttention mechanism was proposed. The model used DeepLabV3 as the backbone network and incorporated the CoTAttention module to enhance the feature extraction capability of the network. To verify the classification performance of the improved network, experiments were conducted using Sentinel-2 remote sensing images from the Tailan River Irrigation District. The results showed that the PointRend model achieved good performance metrics, with pixel accuracy, mean intersection over union (MIoU), and F1 reaching 89%, 88%, and 88%, respectively. Compared with the original PointRend model without the CoTAttention mechanism, the improved PointRend model demonstrated enhanced metrics, with pixel accuracy, MIoU, and F1 increasing by 3, 2, and 3 percentage points, respectively. The improved PointRend model significantly enhanced the ability to capture fine details and refine edges in segmentation tasks, thereby improving overall classification accuracy.

Key words: deep learning, saline-alkali land, remote sensing images, inversion model, attention mechanism

CLC Number:

TP753

BAN Ruo-nan, DONG Luan. Construction of a remote sensing image inversion model for saline-alkali land based on deep learning[J]. HUBEI AGRICULTURAL SCIENCES, 2025, 64(6): 185-189.

References

[1] 吴海华,刘元元,喜炜,等.新疆盐碱耕地改良利用现状与对策建议[J]. 中国农业综合开发, 2023(8):6-8.
[2] 吕宁,石磊,戴昱余,等.新疆盐碱地治理利用研究回顾与启示[J].灌溉排水学报,2024(7):1-13.
[3] 张凤荣. 分区分类因地制宜综合改造利用盐碱地[J].中国土地, 2023(8):30-32.
[4] TZIOLAS N, TSAKIRIDIS N, BEN-DOR E, et al.A memory-based learning approach utilizing combined spectral sources and geographical proximity for improved VIS-NIR-SWIR soil properties estimation[J]. Geoderma,2019,340:11-24.
[5] 王敬哲,丁建丽,葛翔宇,等.基于星地传感技术的土壤盐渍化监测进展与展望[J].遥感学报,2024,28(9):2187-2208.
[6] 谭旺,刘义,董建华,等.基于Sentinel-2卫星影像和土壤变量的盐渍化土壤水溶性盐基离子含量反演[J].中国农村水利水电,2024(7):210-217.
[7] LIU J, ZHANG L, DONG T, et al.The applicability of remote sensing models of soil salinization based on feature space[J].Sustainability, 2021, 13(24):13711.
[8] 刘霞,胡宇,张圣微,等.达拉特旗黄河南岸盐碱化土壤不同含盐量估算模型对比[J].农业机械学报,2024,55(10):360-370.
[9] MULLER S J, VAN NIEKERK A.An evaluation of supervised classifiers for indirectly detecting salt-affected areas at irrigation scheme level[J]. International journal of applied earth observation and geoinformation, 2016, 49: 138-150.
[10] MAHDIANPARI M, SALEHI B, REZAEE M, et al.Very deep convolutional neural networks for complex land cover mapping using multispectral remote sensing imagery[J]. Remote sensing, 2018, 10(7): 1119.
[11] 刘春,杨怿,周源,等.城市河网的低空遥感影像全卷积神经网络水质等级分类[J].同济大学学报(自然科学版),2020,48(3):456-462.
[12] LI Y, YAO T, PAN Y,et al.Contextual transformer networks for visual recognition[J]. IEEE transactions on pattern analysis and machine intelligence,2022, 45(2):1489-1500.
[13] HUANG Y,XU H.Fully convolutional network with attention modules for semantic segmentation[J].Signal image and video processing, 2021,15:1031-1039.
[14] WISAENG K.U-Net++ DSM: Improved U-Net++ for brain tumor segmentation with deep supervision mechanism[J]. IEEE access, 2023, 11:132268-132285.
[15] KARANDASHEV I M, TEPLOV G S,KARMANOV A A,et al.Investigating the efficiency of using U-Net, Erf-Net and DeepLabV3 architectures in inverse lithography-based 90 nm photomask generation[J].Optical memory & neural networks, 2023, 32(4):219-225.
[16] KIRILLOV A, WU Y, HE K,et al.PointRend: Image segmentation as rendering[A].CVF conference on computer vision and pattern recognition (CVPR)[C].IEEE, 2020. 9799-9808.