Effects of image input size and resolution by CNN on the classification accuracy for coniferous forest vegetation in western Sichuan
- Vol. 27, Issue 11, Pages: 2640-2652(2023)
Published： 07 November 2023
DOI： 10.11834/jrs.20221868
扫 描 看 全 文
浏览全部资源
扫码关注微信
- About Journal
- Journal Online
- Special Source
摘要
川西亚高山针叶林位于中国西南地区，受多云、多雨、多雾的影响，难以通过卫星影像进行植被分类的研究。为了解决这一难题，本研究选取川西亚高山针叶林的典型区域王朗自然保护区作为研究区，使用多旋翼无人机获取研究区域北部高分辨率RGB影像，结合卷积神经网络进行植被分类；为进一步挖掘卷积神经网络在无人机遥感影像上的潜力，选择语义分割方法（U-Net）进行分类，并根据不同分辨率的无人机影像和不同尺寸下的样本集构建植被分类模型，建立森林指纹库。结果表明：（1）结合无人机可见光影像和CNN模型进行分类能够获得高精度分类结果。在空间分辨率为5 cm，尺寸为256×256像素的情况下达到最优，总体精度为93.21%，Kappa系数为0.90；（2）选择合适的尺寸大小能够提高模型的分类精度。在5 cm的空间分辨率下，尺寸为128×128像素的模型总体精度为82.30%，Kappa系数为0.76；尺寸为256×256像素的模型总体精度为93.21%，Kappa系数为0.90；（3）超高空间分辨率的升高对模型精度的提升是有限的。当空间分辨率从10 cm升到5 cm时，模型的总体精度提高了0.02，Kappa系数提高了0.03，模型的分类精度并没有明显提升。（4）对于区域内代表性不足的植被类型来说，受空间分辨率和尺寸大小的影响要远高于区域内优势树种，特别是空间分辨率的影响最大。在20 cm的空间分辨率下落叶灌木的生产者精度和用户精度均低于70%。综上，利用无人机高分辨率RGB影像结合CNN模型对川西亚高山针叶林的植被分类能够取得高精度分类结果，本研究可为该区域植被分类提供一种自动、准确的方法。
Abstract
The subalpine coniferous forest in west Sichuan is located in southwest China
which is affected by cloudy
rainy
and foggy conditions. Thus
conducting vegetation classification in the area by using satellite images is difficult.
(Objective)
2
Therefore
this work selects Wanglang Nature Reserve
which is a typical area of subalpine coniferous forest in western Sichuan
as the study area. A multi-rotor UAV is used to acquire high-resolution RGB images of the northern part of the study area
and it is combined with a convolutional neural network model for vegetation classification.
(Method)
2
This study selects the semantic segmentation method (U-Net) for classification
constructs vegetation classification models based on UAV images of different spatial resolutions and sample sets under different tile sizes
and establishes a forest fingerprint library to further explore the potential of convolutional neural networks on UAV remote sensing images.
(Result)
2
(1) The combination of UAV visible images and convolutional neural network model for classification can obtain classification results of high accuracy
which reached the optimum at a spatial resolution of 5 cm and a size of 256×256. The overall accuracy was 93.21%
and the kappa coefficient was 0.90. (2) The increase in ultrahigh spatial resolution had limited improvement on the model accuracy. When the spatial resolution was increased from 10 cm to 5 cm
the overall accuracy of the model improved by 0.02 and the Kappa coefficient improved by 0.03
and the classification accuracy of the model did not improve significantly. (3) Choosing the appropriate size can improve the classification accuracy of the model. Under the spatial resolution of 5 cm
the overall accuracy of the model with the size of 128×128 was 82.30% and the kappa coefficient was 0.76
and the overall accuracy of the model with the size of 256×256 was 93.21% and the kappa coefficient was 0.90. (4) For the vegetation types that were underrepresented in the region
the influence of spatial resolution and tile size was much higher than that of the dominant tree species
especially the influence of spatial resolution was the highest. The producer and user accuracies for deciduous shrubs at a spatial resolution of 20 cm were below 70%.
(Conclusion)
2
This study shows that vegetation classification of subalpine coniferous forests in western Sichuan using UAV high-resolution RGB images combined with convolutional neural networks can achieve high-precision classification results. The effects of UAV spatial resolution and tile sizes on the accuracy of convolutional neural network models are explored
which further details the potential of convolutional neural networks on UAV high-resolution RGB images to provide an automatic and accurate research method for vegetation classification in this region.
关键词
无人机RGB影像卷积神经网络川西亚高山针叶林植被分类输入尺寸空间分辨率
Keywords
UAV RGB imageryConvolutional Neural Network(CNN)subalpine coniferous forest in western Sichuanvegetation classificationtile sizespatial resolution
references
Cao Z. 2016. Present situation and forecast on forest investigation and monitoring by unmanned aerial vehicle. Forestry Construction, (6): 1-5
曹忠. 2016. 低空无人机森林资源调查与监测应用现状与展望. 林业建设, (6): 1-5
Colomina I and Molina P. 2014. Unmanned aerial systems for photogrammetry and remote sensing: a review. ISPRS Journal of Photogrammetry and Remote Sensing, 92: 79-97 [DOI: 10.1016/j.isprsjprs.2014.02.013http://dx.doi.org/10.1016/j.isprsjprs.2014.02.013]
Dai P Q, Ding L X, Liu L J, Dong L F and Huang Y T. 2020. Tree species identification based on FCN using the visible images obtained from an unmanned aerial vehicle. Laser and Optoelectronics Progress, 57(10): 101001
戴鹏钦, 丁丽霞, 刘丽娟, 董落凡, 黄依婷. 2020. 基于FCN的无人机可见光影像树种分类. 激光与光电子学进展, 57(10): 101001 [DOI: 10.3788/LOP57.101001http://dx.doi.org/10.3788/LOP57.101001]
Du Z, Gan S S and Hu J. 2014. Characteristics and protection strategy of forest resources in the alpine region of southwest China. Forest Resources Management, (Z1): 27-31
杜志, 甘世书, 胡觉. 2014. 西南高山林区森林资源特点及保护利用对策探讨. 林业资源管理, (Z1): 27-31 [DOI: 10.13466/j.cnki.lyzygl.2014.S0.007http://dx.doi.org/10.13466/j.cnki.lyzygl.2014.S0.007]
Fassnacht F E, Latifi H, Stereńczak K, Modzelewska A, Lefsky M, Waser L T, Straub C and Ghosh A. 2016. Review of studies on tree species classification from remotely sensed data. Remote Sensing of Environment, 186: 64-87 [DOI: 10.1016/j.rse.2016.08.013http://dx.doi.org/10.1016/j.rse.2016.08.013]
Gu R H. 2005. Analysis and estimation for the landescapes of the nature as the Wanglang national preserve. Journal of Biology, 22(3): 1-4
顾人和. 2005. 王朗自然保护区的大自然景观类型分析与评价. 生物学杂志, 22(3): 1-4 [DOI: 10.3969/j.issn.2095-1736.2005.03.001http://dx.doi.org/10.3969/j.issn.2095-1736.2005.03.001]
Hao Z B, Lin L L, Post C J, Mikhailova E A, Li M H, Chen Y, Yu K Y and Liu J. 2021. Automated tree-crown and height detection in a young forest plantation using mask region-based convolutional neural network (Mask R-CNN). ISPRS Journal of Photogrammetry and Remote Sensing, 178: 112-123 [DOI: 10.1016/j.isprsjprs.2021.06.003http://dx.doi.org/10.1016/j.isprsjprs.2021.06.003]
He C, Zhang S Y and Yao S R. 2014. Forest fires locating technology based on rotor UAV. Bulletin of Surveying and Mapping, (12): 24-27
何诚, 张思玉, 姚树人. 2014. 旋翼无人机林火点定位技术研究. 测绘通报, (12): 24-27 [DOI: 10.13474/j.cnki.11-2246.2014.0389http://dx.doi.org/10.13474/j.cnki.11-2246.2014.0389]
He K M, Zhang X Y, Ren S Q and Sun J. 2016. Deep residual learning for image recognition//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas: IEEE: 770-778 [DOI: 10.1109/CVPR.2016.90http://dx.doi.org/10.1109/CVPR.2016.90]
Huang H Y, Ma X H, Hu L L, Huang Y H and Huang H H. 2021. The preliminary application of the combination of Fast R-CNN deep learning and UAV remote sensing in the monitoring of pine wilt disease. Journal of Environmental Entomology, 43(5): 1295-1303
黄华毅, 马晓航, 扈丽丽, 黄咏槐, 黄焕华. 2021. Fast R-CNN深度学习和无人机遥感相结合在松材线虫病监测中的初步应用研究. 环境昆虫学报, 43(5): 1295-1303 [DOI: 10.3969/j.issn.1674-0858.2021.05.26http://dx.doi.org/10.3969/j.issn.1674-0858.2021.05.26]
Iverson L R, Perera A H, Pastur G M and Peterson U. 2018. Ecosystem services from forest landscapes: where we are and where we go//Perera A H, Peterson U, Martínez Pastur G and Iverson L R, eds. Ecosystem Services from Forest Landscapes. Cham: Springer: 249-258 [DOI: 10.1007/978-3-319-74515-2_9http://dx.doi.org/10.1007/978-3-319-74515-2_9]
Jia L, Zheng K, Tang P and Huo L Z. 2022. Topographic correction effect on classification accuracy for deep neural network classifier—A case study of the U-Net model. National Remote Sensing Bulletin, 26(4): 698-710
贾莉, 郑柯, 唐娉, 霍连志. 2022. 地形校正对U-Net深度神经网络分类器分类精度的影响. 遥感学报, 26(4): 698-710 [DOI: 10.11834/jrs.20229322http://dx.doi.org/10.11834/jrs.20229322]
Kattenborn T, Leitloff J, Schiefer F and Hinz S. 2021. Review on Convolutional Neural Networks (CNN) in vegetation remote sensing. ISPRS Journal of Photogrammetry and Remote Sensing, 173: 24-49 [DOI: 10.1016/j.isprsjprs.2020.12.010http://dx.doi.org/10.1016/j.isprsjprs.2020.12.010]
Liao L L, Zhou L, Wang S Q and Wang X Q. 2016. Carbon sequestration potential of biomass carbon pool for new afforestation in China during 2005-2013. Acta Geographica Sinica, 71(11): 1939-1947
廖亮林, 周蕾, 王绍强, 汪小钦. 2016. 2005-2013年中国新增造林植被生物量碳库固碳潜力分析. 地理学报, 71(11): 1939-1947 [DOI: 10.11821/dlxb201611006http://dx.doi.org/10.11821/dlxb201611006]
Liao X H, Xiao Q and Zhang H. 2019. UAV remote sensing: popularization and expand application development trend. Journal of Remote Sensing, 23(6): 1046-1052
廖小罕, 肖青, 张颢. 2019. 无人机遥感: 大众化与拓展应用发展趋势. 遥感学报, 23(6): 1046-1052 [DOI: 10.11834/jrs.20199422http://dx.doi.org/10.11834/jrs.20199422]
Liao X H, Zhou C H, Su F Z, Lu H Y, Yue H Y and Gou J P. 2016. The mass innovation era of UAV remote sensing. Journal of Geo-Information Science, 18(11): 1439-1447
廖小罕, 周成虎, 苏奋振, 卢海英, 岳焕印, 缑吉平. 2016. 无人机遥感众创时代. 地球信息科学学报, 18(11): 1439-1447 [DOI: 10.3724/SP.J.1047.2016.01439http://dx.doi.org/10.3724/SP.J.1047.2016.01439]
Lin Z W, Ding Q L, Huang J H, Tu W H, Hu D and Liu J F. 2019. Study on tree species classification of UAV optical image based on DenseNet. Remote Sensing Technology and Application, 34(4): 704-711
林志玮, 丁启禄, 黄嘉航, 涂伟豪, 胡典, 刘金福. 2019. 基于DenseNet的无人机光学图像树种分类研究. 遥感技术与应用, 34(4): 704-711 [DOI: 10.11873/j.issn.1004-0323.2019.4.0704http://dx.doi.org/10.11873/j.issn.1004-0323.2019.4.0704]
Liu Q, Wu Y and He H. 2001. Ecological problems of subalpine coniferous forest in the southwest of China. World Sci-Tech R and D, 23(2): 63-69
刘庆, 吴彦, 何海. 2001. 中国西南亚高山针叶林的生态学问题. 世界科技研究与发展, 23(2): 63-69 [DOI: 10.3969/j.issn.1006-6055.2001.02.016http://dx.doi.org/10.3969/j.issn.1006-6055.2001.02.016]
Liu Q W, Li S M, Li Z Y, Fu L Y and Hu K L. 2017. Review on the applications of UAV-Based LiDAR and photogrammetry in forestry. Scientia Silvae Sinicae, 53(7): 134-148
刘清旺, 李世明, 李增元, 符利勇, 胡凯龙. 2017. 无人机激光雷达与摄影测量林业应用研究进展. 林业科学, 53(7): 134-148 [DOI: 10.11707/j.1001-7488.20170714http://dx.doi.org/10.11707/j.1001-7488.20170714]
Maes W H and Steppe K. 2019. Perspectives for remote sensing with unmanned aerial vehicles in precision agriculture. Trends in Plant Science, 24(2): 152-164 [DOI: 10.1016/j.tplants.2018.11.007http://dx.doi.org/10.1016/j.tplants.2018.11.007]
Natesan S, Armenakis C and Vepakomma U. 2019. Resnet-based tree species classification using uav images. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-2-W13: 475-481 [DOI: 10.5194/isprs-archives-XLII-2-W13-475-2019http://dx.doi.org/10.5194/isprs-archives-XLII-2-W13-475-2019]
Pajares G. 2015. Overview and current status of remote sensing applications based on unmanned aerial vehicles (UAVs). Photogrammetric Engineering and Remote Sensing, 81(4): 281-330 [DOI: 10.14358/Pers.81.4.281http://dx.doi.org/10.14358/Pers.81.4.281]
Ronneberger O, Fischer P and Brox T. 2015. U-Net: convolutional networks for biomedical image segmentation//18th International Conference on Medical Image Computing and Computer-Assisted Intervention. Munich: Springer: 234-241 [DOI: 10.1007/978-3-319-24574-4_28http://dx.doi.org/10.1007/978-3-319-24574-4_28]
Schiefer F, Kattenborn T, Frick A, Frey J, Schall P, Koch B and Schmidtlein S. 2020. Mapping forest tree species in high resolution UAV-based RGB-imagery by means of convolutional neural networks. ISPRS Journal of Photogrammetry and Remote Sensing, 170: 205-215 [DOI: 10.1016/j.isprsjprs.2020.10.015http://dx.doi.org/10.1016/j.isprsjprs.2020.10.015]
Trier Ø D, Salberg A B, Kermit M, Rudjord Ø, Gobakken T, Næsset E and Aarsten D. 2018. Tree species classification in Norway from airborne hyperspectral and airborne laser scanning data. European Journal of Remote Sensing, 51(1): 336-351 [DOI: 10.1080/22797254.2018.1434424http://dx.doi.org/10.1080/22797254.2018.1434424]
Wang J, Zhang C, Chen Q, Li H Y, Peng X, Bai M X, Xu Z Y, Liu H D and Chen Y F. 2022. The method of extracting information of cunninghamia lanceolata crown combined with RGB and LiDAR based on UAV. Journal of Southwest Forestry University (Natural Sciences), 42(1): 133-141
王娟, 张超, 陈巧, 李华玉, 彭希, 白明雄, 徐志扬, 刘浩栋, 陈永富. 2022. 结合无人机可见光和激光雷达数据的杉木树冠信息提取. 西南林业大学学报(自然科学), 42(1): 133-141 [DOI: 10.11929/j.swfu.202101013http://dx.doi.org/10.11929/j.swfu.202101013]
Wang X Q, Wang M M, Wang S Q and Wu Y D. 2015. Extraction of vegetation information from visible unmanned aerial vehicle images. Transactions of the Chinese Society of Agricultural Engineering, 31(5): 152-159
汪小钦, 王苗苗, 王绍强, 吴云东. 2015. 基于可见光波段无人机遥感的植被信息提取. 农业工程学报, 31(5): 152-159 [DOI: 10.3969/j.issn.1002-6819.2015.05.022http://dx.doi.org/10.3969/j.issn.1002-6819.2015.05.022]
Wang Y, He C, Liu B L and Li S S. 2022. Single wood parameters extraction and DBH model construction based on UAV tilt photography technology. Journal of Southwest Forestry University, 42(1): 166-173
王越, 何诚, 刘柏良, 李双双. 2022. 基于无人机倾斜摄影技术的单木参数提取及胸径模型构建. 西南林业大学学报, 42(1): 166-173 [DOI: 10.11929/j.swfu.202009047http://dx.doi.org/10.11929/j.swfu.202009047]
Xu Y, Zhang C L, Jiang R J, Wang Z F, Zhu M C and Shen G C. 2021. UAV-based hyperspectral images and monitoring of canopy tree diversity. Biodiversity Science, 29(5): 647-660 [DOI: 10.17520/biods.2021013http://dx.doi.org/10.17520/biods.2021013]
Zhang B, Zhao L and Zhang X L. 2020a. Three-dimensional convolutional neural network model for tree species classification using airborne hyperspectral images. Remote Sensing of Environment, 247: 111938 [DOI: 10.1016/j.rse.2020.111938http://dx.doi.org/10.1016/j.rse.2020.111938]
Zhang C, Atkinson P M, George C, Wen Z F, Diazgranados M and Gerard F. 2020b. Identifying and mapping individual plants in a highly diverse high-elevation ecosystem using UAV imagery and deep learning. ISPRS Journal of Photogrammetry and Remote Sensing, 169: 280-291 [DOI: 10.1016/j.isprsjprs.2020.09.025http://dx.doi.org/10.1016/j.isprsjprs.2020.09.025]
Zhang C, Xia K, Feng H L, Yang Y H and Du X C. 2021. Tree species classification using deep learning and RGB optical images obtained by an unmanned aerial vehicle. Journal of Forestry Research, 32(5): 1879-1888 [DOI: 10.1007/s11676-020-01245-0http://dx.doi.org/10.1007/s11676-020-01245-0]
Zheng J P, Fu H H, Li W J, Wu W Z, Yu L, Yuan S, Tao W Y W, Pang T K and Kanniah K D. 2021. Growing status observation for oil palm trees using Unmanned Aerial Vehicle (UAV) images. ISPRS Journal of Photogrammetry and Remote Sensing, 173: 95-121 [DOI: 10.1016/j.isprsjprs.2021.01.008http://dx.doi.org/10.1016/j.isprsjprs.2021.01.008]
Zhu W Z, Sheng Z L, Shu S M, Wang W W and Li X. 2021. Comprehensive evaluation of ecological function restoration of the natural secondary forests in a subalpine region of western Sichuan. Chinese Journal of Applied and Environmental Biology, 27(3): 694-704
朱万泽, 盛哲良, 舒树淼, 王文武, 李霞. 2021. 川西亚高山天然次生林生态功能恢复综合评价. 应用与环境生物学报, 27(3): 694-704 [DOI: 10.19675/j.cnki.1006-687x.2021.02029http://dx.doi.org/10.19675/j.cnki.1006-687x.2021.02029]
Alert me when the article has been cited
提交
相关文章
相关作者
暂无数据
相关机构
University of Chinese Academy of Sciences, College of Resources and Environment
Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Key Laboratory of Ecosystem Network Observation and Modeling
Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, State Key Laboratory of Resources and Environmental Information System
China University of Geosciences (Wuhan), School of Geography and Information Engineering, Laboratory of Regional Ecological Processes and Environmental Evolution
State Key Laboratory of Earth Surface Processes and Resource Ecology, Faculty of Geographical Science, Beijing Normal University
选择翻译语言
- Postal code：100190
- Tel：010-58887052 Email：nrsb@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备20021838号-9
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0
Bulk Citation
Export to
TOP