Semantic Segmentation and Analysis on Sensitive Parameters of Forest Fire Smoke Using Smoke-Unet and Landsat-8 Imagery
"> Figure 1
<p>Spatial distribution of study regions in the datasets.</p> "> Figure 2
<p>Different intercontinental data distribution.</p> "> Figure 3
<p>Different land cover types of datasets. (<b>a</b>) Ocean; (<b>b</b>) City; (<b>c</b>) Bare soil; (<b>d</b>) Agricultural land; (<b>e</b>) Grassland; (<b>f</b>) Forest. Different intercontinental data distribution.</p> "> Figure 4
<p>Period of fire occurrence.</p> "> Figure 5
<p>The proportion of smoke pixels of different images.</p> "> Figure 6
<p>Smoke-Unet.</p> "> Figure 7
<p>The results of segmentation of different networks. (<b>a</b>) Image acquired over British Columbia, Canada, on 4 August 2017, the smoke is depicted in red line area; (<b>b</b>) The segmentation results of smoke over British Columbia, the smoke pixels are depicted in aqua color; (<b>c</b>) Image acquired over New Zealand area, on 7 Feb 2019, the smoke is depicted in red line area; (<b>d</b>) The segmentation results of smoke over New Zealand area, the smoke pixels are depicted in aqua color.</p> "> Figure 8
<p>The first line shows true-color composition RGB images of smoke plumes. (<b>a1</b>–<b>a14</b>) Siberia area, Russia, on 17 March 2018; (<b>b1</b>–<b>b14</b>) British Columbia, Canada, on 4 August 2017; (<b>c1</b>–<b>c14</b>) Amazon region, Brazil, on 9 August 2019; (<b>d1</b>–<b>d14</b>) New Zealand area, on 7 Feb 2019; (<b>e1</b>–<b>e14</b>) Zambia, on 26 June 2017; (<b>f1</b>–<b>f14</b>) Liangshan region, China, on 21 May 2019. All rows except the first are segmentation results of smoke with different input data, the smoke pixels are depicted in aqua color.</p> "> Figure 8 Cont.
<p>The first line shows true-color composition RGB images of smoke plumes. (<b>a1</b>–<b>a14</b>) Siberia area, Russia, on 17 March 2018; (<b>b1</b>–<b>b14</b>) British Columbia, Canada, on 4 August 2017; (<b>c1</b>–<b>c14</b>) Amazon region, Brazil, on 9 August 2019; (<b>d1</b>–<b>d14</b>) New Zealand area, on 7 Feb 2019; (<b>e1</b>–<b>e14</b>) Zambia, on 26 June 2017; (<b>f1</b>–<b>f14</b>) Liangshan region, China, on 21 May 2019. All rows except the first are segmentation results of smoke with different input data, the smoke pixels are depicted in aqua color.</p> "> Figure 9
<p>The segmentation results of smoke with variety bands combination. (<b>a</b>) The result of Jaccard and Accuracy; (<b>b</b>) The result of recall and F1.</p> "> Figure 10
<p>The image of smoke acquired over British Columbia, Canada, on 4 August 2017. (<b>a</b>) The true-color composition image. (<b>b</b>) The image of smoke after logarithmic transformed. Different targets are marked with numbers 1 through 8. (1) The cloud; (2) The heavy smoke; (3) The thin smoke over area 3; (4) The thin smoke over area 4; (5) The smoke over the hot spot; (6) The soil; (7) The water; (8) The vegetation.</p> "> Figure 11
<p>The spectral profile of different objects. (<b>a</b>) The profile of cloud on area 1; (<b>b</b>) The profile of heavy smoke on area 2; (<b>c</b>) The profile of thin smoke over the area 3; (<b>d</b>) The profile of thin smoke over the area 4; (<b>e</b>) The profile of smoke over the hot spot (the fire point) on area 5.</p> "> Figure 12
<p>The first line is true-color composition RGB images of smoke plumes. (<b>a1</b>–<b>a5</b>) Siberia area, Russia on 17 Mar 2018; (<b>b1</b>–<b>b5</b>) British Columbia, Canada, on 4 August 2017; (<b>c1</b>–<b>c5</b>) Amazon region, Brazil, on 9 August 2019; (<b>d1</b>–<b>d5</b>) New Zealand area, on 7 February 2019; (<b>e1</b>–<b>e5</b>) Zambia, on 26 June 2017; (<b>f1</b>–<b>f5</b>) Liangshan region, China, on 21 May 2019. All rows except the first are segmentation results of smoke with multiple bands and remote sensing indexes, the smoke pixels are depicted in aqua color.</p> ">
Abstract
:1. Introduction
2. Data
2.1. Landsat-8 Multispectral Data
2.2. Study Area
2.3. Fire Seasons
2.4. Proportion of Smoke Pixel
2.5. Training and Validation Dataset
3. Methods
4. Results and Discussion
4.1. Experimental Environment
4.2. Implementation Details
4.3. Implementation Details
4.4. Ablation and Comparative Analysis
4.5. Sensitivity Analysis
4.5.1. Sensitivity of Bands
4.5.2. Sensitivity of Remote Sensing Parameters
5. Conclusions
Author Contributions
Funding
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Sun, G.; Hallema, D.; Asbjornsen, H. Ecohydrological processes and ecosystem services in the Anthropocene: A review. Ecol. Process. 2017, 6, 35. [Google Scholar] [CrossRef] [Green Version]
- Hansen, M.C.; Loveland, T.R. A review of large area monitoring of land cover change using Landsat data. Remote Sens. Environ. 2012, 122, 66–74. [Google Scholar] [CrossRef]
- Houghton, R. Historic role of forests in the global carbon cycle. In Carbon Dioxide Mitigation in Forestry and Wood Industry; Springer: Berlin/Heidelberg, Germany, 1998; pp. 1–24. [Google Scholar]
- Canadell, J.G.; Raupach, M.R. Managing forests for climate change mitigation. Science 2008, 320, 1456–1457. [Google Scholar] [CrossRef] [Green Version]
- Bowman, D.M.J.S.; Balch, J.K.; Artaxo, P.; Bond, W.J.; Carlson, J.M.; Cochrane, M.A.; D’Antonio, C.M.; DeFries, R.S.; Doyle, J.C.; Harrison, S.P.; et al. Fire in the earth system. Science 2009, 324, 481–484. [Google Scholar] [CrossRef]
- Allen, C.D.; Macalady, A.K.; Chenchouni, H.; Bachelet, D.; McDowell, N.; Vennetier, M.; Kitzberger, T.; Rigling, A.; Breshears, D.D.; Hogg, E.H.T. A global overview of drought and heat-induced tree mortality reveals emerging climate change risk for forests. For. Ecol. Manag. 2010, 259, 660–684. [Google Scholar] [CrossRef] [Green Version]
- Hirsch, K.G.; Corey, P.N.; Martell, D.L. Using expert judgment to model initial attack fire crew effectiveness. For. Sci. 1998, 44, 539–549. [Google Scholar]
- Korontzi, S.; McCarty, J.; Loboda, T.; Kumar, S.; Justice, C. Global distribution of agricultural fires in croplands from 3 years of Moderate Resolution Imaging Spectroradiometer (MODIS) data. Glob. Biogeochem. Cycles 2006, 20, GB2021. [Google Scholar] [CrossRef]
- Dwyer, E.; Pinnock, S.; Gregoire, J.M.; Pereira, J.M.C. Global spatial and temporal distribution of vegetation fire as determined from satellite observations. Int. J. Remote Sens. 2000, 21, 1289–1302. [Google Scholar] [CrossRef]
- Csiszar, I.; Denis, L.; Giglio, L.; Justice, C.O.; Hewson, J. Global fire activity from two years of MODIS data. Int. J. Wildland Fire 2005, 14, 117–130. [Google Scholar] [CrossRef]
- Dozier, J. A method for satellite identification of surface temperature fields of subpixel resolution. Remote Sens. Environ. 1981, 11, 221–229. [Google Scholar] [CrossRef]
- Matson, M.; Stephens, G.; Robinson, J. Fire detection using data from the NOAA-N satellites. Int. J. Remote Sens. 1987, 8, 961. [Google Scholar] [CrossRef]
- Robinson, J.M. Fire from space: Global evaluation using infrared remote sensing. Int. J. Remote Sens. 1991, 12, 3–24. [Google Scholar] [CrossRef]
- Kaufman, Y.J.; Tanré, D. Algorithm for Remote Sensing of Tropospheric Aerosol From MODIS. In NASA MODIS Algorithm Theoretical Basis Document; Goddard Space Flight Center: Greenbelt, MD, USA, 1998; Volume 85, pp. 3–68. [Google Scholar]
- Giglio, L.; Descloitres, J.; Justice, C.O.; Kaufman, Y.J. An enhanced contextual fire detection algorithm for MODIS. Remote Sens. Environ. 2003, 87, 273–282. [Google Scholar] [CrossRef]
- Giglio, L.; Schroeder, W.; Justice, C.O. The collection 6 MODIS active fire detection algorithm and fire products. Remote Sens. Environ. 2016, 178, 31–41. [Google Scholar] [CrossRef] [Green Version]
- Justice, C.O.; Román, M.O.; Csiszar, I.; Vermote, E.F.; Wolfe, R.E.; Hook, S.J. Land and cryosphere products fromSuomi NPP VIIRS: Overview and status. J. Geophys. Res. 2013, 118, 9753–9765. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Wolfe, R.E.; Lin, G.; Nishihama, M.; Tewari, K.P.; Tilton, J.C.; Isaacman, A.R. Suomi NPP VIIRS prelaunch and on-orbit geometric calibration and characterization. J. Geophys. Res. 2013, 118, 508–511, 521. [Google Scholar] [CrossRef] [Green Version]
- Csiszar, I.; Schroeder, W.; Giglio, L.; Ellicott, E.; Vadrevu, K.P.; Justice, C.O. Active fires from the Suomi NPP Visible Infrared Imaging Radiometer Suite: Product status and first evaluation results. J. Geophys. Res. 2014, 119, 803–816. [Google Scholar] [CrossRef]
- Schroeder, W.; Oliva, P.; Giglio, L.; Quaryle, B.; Lorenz, E.; Morelli, F. Active fire detection using Landsat-8/OLI data. Remote Sens. Environ. 2015, 185, 210–220. [Google Scholar] [CrossRef] [Green Version]
- Li, Z.; Nadon, S.; Cihlar, J. Satellite-based detection of Canadian boreal forest fires:development and application of the algorithm. Int. J. Remote Sens. 2000, 21, 3057–3069. [Google Scholar] [CrossRef] [Green Version]
- Li, Z.; Kaufman, Y.J.; Ichoku, C.; Fraser, R.; Trishchenko, A.; Giglio, L.; Yu, X. A Review of AVHRR-based Active Fire Detection Algorithms: Principles, Limitations, and Recommendations. 2000. Available online: http://www.fao.org/GTOS/gofc-gold/docs/fire_ov.pdf (accessed on 29 September 2021).
- Csiszar, I.A.; Morisette, J.T.; Giglio, L. Validation of active fire detection from moderateresolution satellite sensors: The MODIS example in northern Eurasia. Remote Sens. 2006, 44, 1757–1764. [Google Scholar] [CrossRef]
- Genet, H.; McGuire, A.D.; Barrett, K.; Breen, A.; Euskirchen, E.S.; Johnstone, J.F.; Yuan, F. Modeling the effects of fire severity and climate warming on active layer thickness and soil carbon storage of black spruce forests across the landscape in interior Alaska. Environ. Res. Lett. 2013, 8, 045016. [Google Scholar] [CrossRef]
- Giglio, L.; Kendall, J.D.; Justice, C.O. Evaluation of global fire detection algorithms using simulated AVHRR infrared data. Int. J. Remote Sens. 1999, 20, 1947–1985. [Google Scholar] [CrossRef]
- Schroeder, W.; Oliva, P.; Giglio, L.; Ivan, A.C. The New VIIRS375m active fire detection data product: Algorithm description and initial assessment. Remote Sens. Environ. 2014, 143, 85–96. [Google Scholar] [CrossRef]
- Waigl, C.F.; Stuefer, M.; Prakash, A.; Ichoku, C. Detecting high and low-intensity fires in Alaska using VIIRS I-band data: An improved operational approach for high latitudes. Remote Sens. Environ. 2017, 199, 389–400. [Google Scholar] [CrossRef]
- Elvidge, C.D.; Zhizhin, M.; Hsu, F.C.; Baugh, K.E. VIIRS Nightfire: Satellite Pyrometry at Night. Remote Sens. 2013, 5, 4423–4449. [Google Scholar] [CrossRef] [Green Version]
- Cho, K.; Kim, Y.; Kim, Y. Disaggregation of Landsat-8 Thermal Data Using Guided SWIR Imagery on the Scene of a Wildfire. Remote Sens. 2018, 10, 105. [Google Scholar] [CrossRef] [Green Version]
- Li, Z.; Khananian, A.; Fraser, R.H.; Cihlar, J. Automatic detection of fire smoke using artificial neural networks and threshold approaches applied to AVHRR imagery. IEEE Trans. Geosci. Remote Sens. 2001, 39, 1859–1870. [Google Scholar]
- Garay, M.J.; Mazzoni, D.M.; Davies, R.; Diner, D. The application of support vector machines to the analysis of global datasets from MISR. In Proceedings of the Fourth Conference on Artificial Intelligence Applications to Environmental Science, San Diego, CA, USA, 9–13 January 2005. [Google Scholar]
- Mazzoni, D.; Garay, M.J.; Davies, R.; Nelson, D. An operational MISR pixel classifier using support vector machines. Remote Sens. Environ. 2006, 107, 149–158. [Google Scholar] [CrossRef]
- Mazzoni, D.; Logan, J.A.; Diner, D.; Kahn, R.; Tong, L.; Li, Q. A data-mining approach to associating MISR smoke plume heights with MODIS fire measurements. Remote Sens. Environ. 2007, 107, 138–148. [Google Scholar] [CrossRef]
- Li, X.L.; Wang, J.; Song, W.G.; Ma, J.; Telesca, L.; Zhang, Y.M. Automatic Smoke Detection in MODIS Satellites Data based on K-means Clustering and Fisher Linear Discrimination. Photogramm. Eng. Remote Sens. 2014, 80, 971–982. [Google Scholar] [CrossRef]
- Li, X.L.; Song, W.G.; Lian, L.; Wei, X. Forest Fire Smoke Detection Using Back-Propagation Neural Network Based on MODIS Data. Remote Sens. 2015, 7, 4473–4498. [Google Scholar] [CrossRef] [Green Version]
- Zhang, Z.; Liu, Q.; Wang, Y. Road extraction by deep residual u-net. IEEE Geosci. Remote Sens. Lett. 2018, 15, 749–753. [Google Scholar] [CrossRef] [Green Version]
- Jeppesena, J.H.; Jacobsena, R.H. A cloud detection method for landsat 8 images based on pcanet. Remote Sens. 2018, 10, 877. [Google Scholar]
- Ba, R.; Chen, C.; Yuan, J.; Song, W.; Lo, S. SmokeNet:Satellites Smoke Scene Detection Using Convolutional Neural Network with Spatial and Channel-Wise Attetion. Remote Sens. 2019, 11, 1702. [Google Scholar] [CrossRef] [Green Version]
- Ronneberger, O.; Fischer, P.; Brox, T. U-Net: Convolutional Networks for Biomedical Image Segmentation. 2015. Available online: https://arxiv.org/pdf/1505.04597.pdf (accessed on 29 September 2021).
- Bao, Y.; Liu, W.; Gao, O.; Lin, Z.; Hu, Q. E-Unet++: A Semantic Segmentation Method for Remote Sensing Images. In Proceedings of the 2021 IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC), Chongqing, China, 18–20 June 2021; pp. 1858–1862. [Google Scholar] [CrossRef]
- Li, X.; Du, Z.; Huang, Y.; Tan, Z. A deep translation (GAN) based change detection network for optical and SAR remote sensing images. ISPRS J. Photogramm. Remote Sens. 2021, 179, 14–34. [Google Scholar] [CrossRef]
- Maratkhan, A.; Ilyassov, I.; Aitzhanov, M.; Demirci, M.F.; Ozbayoglu, A.M. Deep learning-based investment strategy: Technical indicator clustering and residual blocks. Soft Comput. 2021, 25, 5151–5161. [Google Scholar] [CrossRef]
- Kastner, S.; Ungerleider, L.G. Mechanisms of visual attention in the human cortex. Annu. Rev. Neurosci. 2000, 23, 315–341. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Balshi, M.S.; McGuire, A.D.; Zhuang, Q.; Melillo, J.; Kicklighter, D.W.; Kasischke, E. The role of historical fire disturbance in the carbon dynamics of the pan-boreal region: A process-based analysis. J. Geophys. Res. 2007, 112, 1–18. [Google Scholar] [CrossRef] [Green Version]
- Dennison, P.E.; Brewer, S.C.; Arnold, J.D.; Moritz, M.A. Large wildfire trends in the western United States, 1984–2011. Geophys. Res. Lett. 2014, 41, 2928–2933. [Google Scholar] [CrossRef]
- Shelhamer, E.; Long, J.; Darrell, T. Fully Convolutional Networks for Semantic Segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 39, 640–651. [Google Scholar] [CrossRef]
- Badrinarayanan, V.; Kendall, A.; Cipolla, R. SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 39, 2481–2495. [Google Scholar] [CrossRef] [PubMed]
- Zhao, H.; Shi, J.; Qi, X.; Wang, X.; Jia, J. Pyramid Scene Parsing Network. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 21–26 July 2017; pp. 6230–6239. [Google Scholar]
Payload Name | Band Number | Band Name | Spectral Range(nm) | Resolution(m) |
---|---|---|---|---|
OLI | 1 | Coastal | 433~453 | 30 |
2 | Blue | 450~515 | 30 | |
3 | Green | 525~600 | 30 | |
4 | Red | 630~680 | 30 | |
5 | NIR | 845~885 | 30 | |
6 | SWIR1 | 560~660 | 30 | |
7 | SWIR2 | 100~300 | 30 | |
8 | Panchromatic | 500~680 | 15 | |
9 | Cirrus | 1360~1390 | 30 | |
TIRS | 10 | TIRS1 | 1060~1119 | 60 |
11 | TIRS2 | 1150~1251 | 60 |
Programming Environment | Auxiliary Library | Hardware Configuration | Other Software |
---|---|---|---|
Python3.5 | Shapely | CPU:[email protected] GHz | |
Tensorflow1.9 | Opencv2.2 | GPU:NVDIA TITAN X | ENVI5.3 |
CUDA8.0 | Tifffile0.12 | RAM:16 GB | ArcGIS10.3 |
cuDNN10.0 | Rasterio1.1.2 | Numba0.26.0 | Scikit_image0.12.3 |
Keras2.2.0 | h5py2.6.0 |
Number | Data Type | Item | Band |
---|---|---|---|
1 | Band Data | Multispectral Band | 1–7, 10 |
2 | Band Data | Panchromatic Band | 8 |
3 | Remote Sensing Index | EVI | / |
4 | Remote Sensing Index | NBR | / |
5 | Remote Sensing Index | AOD | / |
6 | Remote Sensing Index | BT | / |
Network | Dataset | Loss | Jaccard | Accuracy | Recall | F1 |
---|---|---|---|---|---|---|
Unet | Train | 0.844 | 0.657 | 0.801 | 0.753 | 0.773 |
Validation | 1.889 | 0.699 | 0.694 | 0.781 | 0.735 | |
Res-Unet | Train | 0.851 | 0.690 | 0.805 | 0.829 | 0.813 |
Validation | 1.636 | 0.59 | 0.701 | 0.944 | 0.805 | |
Atten-Res-Unet | Train | 1.514 | 0.703 | 0.835 | 0.816 | 0.823 |
Validation | 1.926 | 0.654 | 0.696 | 0.894 | 0.782 | |
FCN | Train | 1.479 | 0.735 | 0.845 | 0.852 | 0.844 |
Validation | 1.974 | 0.58 | 0.711 | 0.811 | 0.758 | |
Segnet | Train | 1.532 | 0.712 | 0.831 | 0.835 | 0.828 |
Validation | 1.708 | 0.665 | 0.761 | 0.841 | 0.799 | |
PSPnet | Train | 1.406 | 0.748 | 0.845 | 0.871 | 0.851 |
Validation | 1.901 | 0.581 | 0.751 | 0.812 | 0.765 | |
Smoke-Unet | Train | 0.759 | 0.752 | 0.923 | 0.917 | 0.918 |
Validation | 1.134 | 0.644 | 0.725 | 0.838 | 0.775 |
Number | Data Type | Data Dimension | Band |
---|---|---|---|
1 | RGB | 3 | Band 2~4 |
2 | RGB + NIR | 4 | Band 2~5 |
3 | RGB + TIRS1 | 4 | Band 2~4,10 |
4 | RGB + SWIR2 | 4 | Band 2~4,7 |
5 | RGB + SWIR1 + SWIR2 | 5 | Band 2~4,6,7 |
6 | RGB + SWIR1 + NIR | 5 | Band 2~6 |
7 | RGB + TIRS1 + SWIR2 | 5 | Band 2~4,7,10 |
8 | TIRS1 | 1 | Band 10 |
9 | NIR + SWIR1/2 + TIRS1 | 4 | Band 5~7,10 |
10 | SWIR1 + NIR + Blue | 3 | Band 2,5,6 |
11 | Multiple | 8 | Band 1~7, Band 10 |
12 | Multiple + Pan | 9 | Band 1~7, Band 10~11 |
13 | All data | 11 | Band 1~11 |
Number | Data Type | Jaccard | Accuracy | Recall | F1 |
---|---|---|---|---|---|
1 | RGB | 0.692 | 0.701 | 0.980 | 0.818 |
2 | RGB + NIR | 0.623 | 0.809 | 0.730 | 0.767 |
3 | RGB + TIRS1 | 0.535 | 0.653 | 0.747 | 0.697 |
4 | RGB + SWIR2 | 0.748 | 0.759 | 0.982 | 0.856 |
5 | RGB + SWIR1 + SWIR2 | 0.701 | 0.707 | 0.988 | 0.824 |
6 | RGB + SWIR1 + NIR | 0.737 | 0.753 | 0.970 | 0.848 |
7 | RGB + TIRS1 + SWIR2 | 0.700 | 0.709 | 0.981 | 0.823 |
8 | TIRS1 | 0.294 | 0.585 | 0.371 | 0.455 |
9 | NIR + SWIR1/2 + TIRS1 | 0.479 | 0.852 | 0.522 | 0.648 |
10 | SWIR1 + NIR + Blue | 0.305 | 0.322 | 0.855 | 0.468 |
11 | Multiple | 0.646 | 0.658 | 0.814 | 0.784 |
12 | Multiple + Pan | 0.673 | 0.701 | 0.844 | 0.804 |
13 | All data | 0.683 | 0.801 | 0.825 | 0.809 |
Number | Data Type | Data Dimension |
---|---|---|
1 | RGB + SWIR2 + EVI | 5 |
2 | RGB + SWIR2 + NBR | 5 |
3 | RGB + SWIR2 + BT | 5 |
4 | RGB + SWIR2 + AOD | 5 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Wang, Z.; Yang, P.; Liang, H.; Zheng, C.; Yin, J.; Tian, Y.; Cui, W. Semantic Segmentation and Analysis on Sensitive Parameters of Forest Fire Smoke Using Smoke-Unet and Landsat-8 Imagery. Remote Sens. 2022, 14, 45. https://doi.org/10.3390/rs14010045
Wang Z, Yang P, Liang H, Zheng C, Yin J, Tian Y, Cui W. Semantic Segmentation and Analysis on Sensitive Parameters of Forest Fire Smoke Using Smoke-Unet and Landsat-8 Imagery. Remote Sensing. 2022; 14(1):45. https://doi.org/10.3390/rs14010045
Chicago/Turabian StyleWang, Zewei, Pengfei Yang, Haotian Liang, Change Zheng, Jiyan Yin, Ye Tian, and Wenbin Cui. 2022. "Semantic Segmentation and Analysis on Sensitive Parameters of Forest Fire Smoke Using Smoke-Unet and Landsat-8 Imagery" Remote Sensing 14, no. 1: 45. https://doi.org/10.3390/rs14010045