Influence of U-Net model on the accuracy of shelter forest extraction with different spatial resolutions

doi:10.13474/j.cnki.11-2246.2021.0173

Abstract

Abstract: Aiming at the problem of high-accuracy of UAV image acquisition but small data scale, this paper proposes to use U-Net model to explore the influence of different spatial resolution on the extraction accuracy of farmland shelterbelts. Take the CW-20 compound-wing UAV equipped with Micro MAC12 Snap multispectral sensor obtained three different height of 300 m(spatial resolution 0.15 m), 400 m(spatial resolution 0.20 m) and 500 m(spatial resolution 0.25 m) as an example. For remote sensing images, experiments results have shown that the accuracy error of image extraction at three different height is within 1.3%, and the accuracy error of MIoU is within 3.7%. The spatial resolution had little effect on the extraction accuracy of shelterbelts. The image data with high spatial resolution can not significantly improve the extraction accuracy of shelterbelts. This study provides a theoretical basis for the acquisition of large-scale agricultural and forestry remote sensing monitoring data sources.

Key words: UAV, multi-spectral image, spatial resolution, shelterbelts extraction, U-Net model

CLC Number:

P237

WANG Xuewen, ZHAO Qingzhan, TIAN Wenzhong, LONG Xiang, JIANG Ping. Influence of U-Net model on the accuracy of shelter forest extraction with different spatial resolutions[J]. Bulletin of Surveying and Mapping, 2021, 0(6): 39-43.

References

[1] HU Y F, ZHANG Q L, ZHANG Y, et al. A deep convolu-tion neural network method for land cover mapping: a case study of Qinhuangdao, China[J]. Remote Sensing, 2018, 10(12):2053.
[2] 张磊,宫兆宁,王启为,等.Sentinel-2影像多特征优选的黄河三角洲湿地信息提取[J].遥感学报,2019,23(2):313-326.
[3] DASH J P, WATT M S, PEARSE G D, et al. Assessing very high resolution UAV imagery for monitoring forest health during a simulated disease outbreak[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2017, 131:1-14.
[4] DENG L, MAO Z H, LI X J, et al. UAV-based multi-spectral remote sensing for precision agriculture: a comparison between different cameras[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2018, 146:124-136.
[5] 戴建国,张国顺,郭鹏,等. 基于无人机遥感可见光影像的北疆主要农作物分类方法[J]. 农业工程学报,2018,34(18):122-129.
[6] TU Y H, JOHANSEN K, PHINN S, et al. Measuring canopy structure and condition using multi-spectral UAS imagery in a horticultural environment[J]. Remote Sensing, 2019, 11(3):269.
[7] 刘舒,朱航. 基于超高空间分辩率无人机影像的面向对象土地利用分类方法[J]. 农业工程学报,2020,36(2):87-94.
[8] FISHER J R B, PORRAS E A A, DENNEDY-FRANK P J, et al. Impact of satellite imagery spatial resolution on land use classification accuracy and modeled water quality[J].Remote Sensing in Ecology and Conservation, 2017,4(2):137-149.
[9] LI N, XIE G D, ZHOU D M, et al. Remote sensing classification of marsh wetland with different resolution images[J]. Journal of Resources and Ecology, 2016, 7(2):107-114.
[10] YE H C, HUANG W J, HUANG S Y, et al. Recognition of banana fusarium wilt based on UAV remote sensing[J].Remote Sensing, 2020,12(6):938.
[11] 邱琳,卢必慧,孙玲,等.GF-1卫星影像的空间分辨率对水稻识别精度的影响[J].江苏农业学报,2019,35(1):70-75.
[12] 刘婷,陈晨,范文义,等.基于不同空间尺度遥感影像估算森林叶面积指数的差异[J].应用生态学报,2019,30(5):1687-1698.
[13] RONNEBERGER O, FISCHER P, BROX T. U-Net: convo-lutional networks for biomedical image segmentation[J]. Computer Vision and Pattern Recongnition, 2015: 1505.04597.
[14] MAGGIORI E,TARABALKA Y, CHARPIAT G, et al. Fully convolutional neural networks for remote sensing image classification[C]// IEEE International Geoscience and Remote Sensing Symposium. Beijing: IEEE, 2016.
[15] 许慧敏. 基于深度学习U-Net模型的高分辨率遥感影像分类方法研究[D]. 成都:西南交通大学,2018.
[16] 周志华.机器学习[M].北京:清华大学出版社,2016.
[17] HAMDI Z M, BRANDMEIER M, STRAUB C. Forest damage assessment using deep learning on high resolution remote sensing data[J]. Remote Sensing, 2019,11(17):1976.