Précédent|Prochain

Forest burned area detection with time series data based on Stacked ConvLSTM

  • role： First author第一作者
  • Affiliation：

    Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China

    University of Chinese Academy of Sciences, Beijing 100049, China

  • Email：lishujun@aircas.ac.cn
  • Introduction：李淑君，研究方向为遥感图像处理。E-mail： lishujun@aircas.ac.cn
LI Shujun12,  
  • Affiliation：

    Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China

ZHENG Ke1,  
  • Affiliation：

    Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China

TANG Ping1,  
  • role： Corresponding author通信作者
  • Affiliation：

    Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, China

  • Email：huolz@aircas.ac.cn
  • Introduction：霍连志，研究方向为遥感图像处理和分类。E-mail： huolz@aircas.ac.cn
HUO Lianzhi1*,  
  • Affiliation：

    Nanjing University of Posts and Telecommunications, Nanjing 210042, China

YUAN Yuan3

résumé

As the largest land cover, forests play an important role in human living environment, biological habitat, and global carbon cycle. Forest health is directly related to global ecological security and sustainable development of human society. In recent years, urban construction, disasters, forest management and deforestation, and other factors have caused different degrees of disturbance to forests. It is important to determine the exact time point and spatial range of forest burned area for forest damage assessment, management, carbon accounting, and forest restoration management. Owing to the continuity of spatial distribution of forest burned areas, most of the existing methods of forest burned area extraction use the two-step treatment strategy of first classification and then post-processing to suppress the effect of false alarm pixels. In this paper, a spatiotemporal detection method, Stacked ConvLSTM, is proposed for the detection of forest fire tracks in time series. This method avoids subjective post-processing operations on the basis of maintaining better spatial continuity of the results, and achieves end-to-end extraction of forest burned area information, which improves the extraction accuracy of forest fire-burning land. This paper proposes to use Stacked ConvLSTM to detect forest disturbance in time and space. Combined with the characteristics of ConvLSTM in extracting temporal and spatial characteristics from long-term historical series, it can predict the change trend of vegetation in a period of time in the future, and accurately determine the time point and spatial range of forest disturbance. ConvLSTM is an LSTM variant proposed on the basis of LSTM. The full connection state from input layer to hidden layer and from hidden layer to hidden layer of LSTM is replaced by convolution connection, which can make full use of spatial information. Compared with single-pixel-based methods, ConvLSTM can extract the spatiotemporal structure information of time series images at the same time, which is better for spatiotemporal analysis. In this paper, Stacked ConvLSTM is used to detect the temporal and spatial distribution of forest burned areas, predict the change trend of vegetation in a period of time in the future, and determine the presence of forest burned areas by comparing with the newest time-series images. With MODIS long time series data, based on the historical time series of Yinanhe Forest Farm of Zhanhe Forestry Bureau in Heilongjiang Province and Beidahe Forest Farm of Bilahe Forestry Bureau in Inner Mongolia from 2001—2008 and 2001—2016, the extraction results of burned areas were compared with Stacked LSTM and bfast algorithm. The Stacked ConvLSTM, Stacked LSTM, and bfast algorithms were used to extract forest burned areas from MODIS time series in both regions, and to compare the detection results with the Fire_CCI 5.1 burned areas products released by ESA. Results show that, firstly, from the visual effect, in study area Ⅰ, the error detection of Stacked ConvLSTM is fewer than that of Stacked LSTM and bfast algorithm and maintains high continuity in spatial distribution. In study Area Ⅱ, Stacked ConvLSTM detected a more complete area of fire. Secondly, in study area Ⅰ , Stacked ConvLSTM was 0.120 and 0.405 more accurate than Stacked LSTM and bfast algorithms, respectively. Moreover, the recall rate, accuracy, and Fire_CCI 5.1 F1-score were higher. In study area Ⅰ , the accuracy of Stacked ConvLSTM is 0.924 had a higher recall rate, accuracy, and F1-score than Stacked LSTM, bfast algorithms, and Fire_CCI 5.1. The detection accuracy of ConvLSTM model in space is higher than that of the other two methods, and its continuity of detection results in space is better. The detection effect of ConvLSTM model is equivalent to that of Stacked LSTM in time, but both of them are closer to the real fire time point than bfast algorithm. Results show that Stacked ConvLSTM has advantages in obtaining the change trend of forest long-term historical series for spatiotemporal prediction, and improves the detection accuracy of forest fire to a certain extent.

mots-clés

Stacked ConvLSTM;time series;spatiotemporal prediction;forest burned area

References

  1. 1.
    Bolton D K, Coops N C and Wulder M A. 2015. Characterizing residual structure and forest recovery following high-severity fire in the western boreal of Canada using Landsat time-series and airborne Lidar data. Remote Sensing of Environment, 163: 48-60
  2. 2.
    Cao Y C, Wu Z P, Zhou Y F, Wei S J and Feng X Q. 2020. Forest fire recognition method based on recurrent neural network. Forestry and Environmental Science, 36(5): 34-40
  3. 3.
    Chance C M, Hermosilla T, Coops N C, Wulder M A, White J C. 2016. Effect of topographic correction on forest change detection using spectral trend analysis of Landsat pixel-based composites. International Journal of Applied Earth Observation and Geoinformation, 44:186–194
  4. 4.
    Cheng M, Xu Q, Lv J M, Liu W Y, Li Q and Wang J P. 2016. MS-LSTM: a multi-scale LSTM model for BGP anomaly detection//2016 IEEE 24th International Conference on Network Protocols (ICNP). Singapore: IEEE: 1-6
  5. 5.
    Chollet F,others.2015.Keras.GitHub.[]
  6. 6.
    Chuvieco E, Lizundia-Loiola J, Pettinari M L, Ramo R, Padilla M, Tansey K, Mouillot F, Laurent P, Storm T, Heil A and Plummer S. 2018. Generation and analysis of a new global burned area product based on MODIS 250 m reflectance bands and thermal anomalies. Earth System Science Data, 10(4): 2015-2031
  7. 7.
    Chuvieco E, Yue C, Heil A, Mouillot F, Alonso-Canas I, Padilla M, Pereira J M, Oom D and Tansey K. 2016. A new global burned area product for climate assessment of fire impacts. Global Ecology and Biogeography, 25(5): 619-629
  8. 8.
    Gers F A and Schmidhuber J. 2000. Recurrent nets that time and count.//Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks//IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium. Como: IEEE: 189-194
  9. 9.
    Hilker T, Wulder M A, Coops N C, Linke J, McDermid G, Masek J G, Gao F and White J C. 2009. A new data fusion model for high spatial- and temporal-resolution mapping of forest disturbance based on Landsat and MODIS. Remote Sensing of Environment, 113(8): 1613-1627
  10. 10.
    Hochreiter S and Schmidhuber J. 1997. Long short-term memory. Neural Computation, 9(8): 1735-1780
  11. 11.
    Huang C Q, Goward S N, Masek J G, Thomas N, Zhu Z L and Vogelmann J E. 2010. An automated approach for reconstructing recent forest disturbance history using dense Landsat time series stacks. Remote Sensing of Environment, 114(1): 183-198
  12. 12.
    Huang C Q, Goward S N, Schleeweis K, Thomas N, Masek J G and Zhu Z L. 2009. Dynamics of national forests assessed using the Landsat record: case studies in eastern United States. Remote Sensing of Environment, 113(7): 1430-1442
  13. 13.
    Huete A, Didan K, Miura T, Rodriguez E P, Gao X and Ferreira L G. 2002. Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sensing of Environment, 83(1/2): 195-213
  14. 14.
    Huo L Z, Tang P, Zhang Z and Tuia D. 2015. Semisupervised classification of remote sensing images with hierarchical spatial similarity. IEEE Geoscience and Remote Sensing Letters, 12(1): 150-154
  15. 15.
    Jordan M I. 1997. Serial order: a parallel distributed processing approach. Advances in Psychology, 121: 471-495
  16. 16.
    Kennedy R E, Cohen W B and Schroeder T A. 2007. Trajectory-based change detection for automated characterization of forest disturbance dynamics. Remote Sensing of Environment, 110(3): 370-386
  17. 17.
    Kennedy R E, Yang Z Q and Cohen W B. 2010. Detecting trends in forest disturbance and recovery using yearly Landsat time series: 1. LandTrendr-temporal segmentation algorithms. Remote Sensing of Environment, 114(12): 2897-2910
  18. 18.
    Khan S H, He X M, Porikli F and Bennamoun M. 2017. Forest change detection in incomplete satellite images with deep neural networks. IEEE Transactions on Geoscience and Remote Sensing, 55(9): 5407-5423
  19. 19.
    Kim S, Hong S, Joh M and Song S K. 2017. DeepRain: ConvLSTM network for precipitation prediction using multichannel radar data. arXiv Preprint arXiv: 1711.02316
  20. 20.
    Kong Y L, Huang Q Q, Wang C Y, Chen J B, Chen J S and He D X. 2018. Long short-term memory neural networks for online disturbance detection in satellite image time series. Remote Sensing, 10(3): 452
  21. 21.
    Längkvist M, Karlsson L and Loutfi A. 2014. A review of unsupervised feature learning and deep learning for time-series modeling. Pattern Recognition Letters, 42: 11-24
  22. 22.
    Liu C, Frazier P, Kumar L.2007. Comparative assessment of the measures of thematic classification accuracy. Remote Sensing of Environment, 107(4): 606–616
  23. 23.
    Markham B L, Helder D L.2012. Forty-year calibrated record of earth-reflected radiance from Landsat: a review. Remote Sensing of Environment, 122:30–40
  24. 24.
    Peng D L, Wu C Y, Li C J, Zhang X Y, Liu Z J, Ye H C, Luo S Z, Liu X J, Hu Y and Fang B. 2017. Spring green-up phenology products derived from MODIS NDVI and EVI: intercomparison, interpretation and validation using National Phenology Network and AmeriFlux observations. Ecological Indicators, 77: 323-336
  25. 25.
    Reddy D S and Prasad P R C. 2018. Prediction of vegetation dynamics using NDVI time series data and LSTM. Modeling Earth Systems and Environment, 4(1): 409-419
  26. 26.
    Ruiz J A M, Lázaro J R G, Cano I D Á and Leal P H. 2014. Burned area mapping in the North American boreal forest using Terra-MODIS LTDR (2001—2011): a comparison with the MCD45A1, MCD64A1 and BAGEOLAND-2 products. Remote Sensing, 6(1): 815-840
  27. 27.
    Shi X J, Chen Z R, Wang H, Yeung D Y, Wong W K and Woo W C. 2015. Convolutional LSTM network: a machine learning approach for precipitation nowcasting//Proceedings of the 28th International Conference on Neural Information Processing Systems. Montreal: MIT Press: 802-810
  28. 28.
    Sun H B, Wang Z L, Zhu Z W, Huang F F, Gan X H and Zhang W Q. 2018. Stoichiometry characteristics of dominant tree species and soil in Heritiera littoralis community in Baguang Wetland Park in Shenzhen. Forestry and Environmental Science, 34(6): 15-20
  29. 29.
    Sun G F，Qin X L，Liu S C，et al. Potential analysis of typical vegetation index for identifying burned area[J].Remote Sensing for Land and Resources，2019，31( 1) : 204-211
  30. 30.
    Verbesselt J, Zeileis A and Herold M. 2012. Near real-time disturbance detection using satellite image time series. Remote Sensing of Environment, 123: 98-108
  31. 31.
    Wang Z S, Wei S J, Xie J H, Zhong Y X, Li X C, Tang H H, Wu Z P, Zhou Y F and Li Q. 2019. Effect of forest fire combustion environment on forest fire suppression. Forestry and Environmental Science, 35(2): 84-88
  32. 32.
    Wang Z X, Liu C and Huete A. 2003. From AVHRR-NDVI to MODIS-EVI: advances in vegetation index research. Acta Ecologica Sinica, 23(5): 979-987
  33. 33.
    Wei S J, Luo S S, Luo B Z, Li X C, Wang Z S, Wu Z P, Zhou Y F, Zhong Y X and Li Q. 2020. Occurrence regularity of forest fire under the background of climate change. Forestry and Environmental Science, 36(2): 133-143
  34. 34.
    Wu J W, Sun L Y, Ji R P, Feng R, Yu W Y and Zhang Y S. 2020. Research progress and prospect of remote sensing on extracting burned areas information. Journal of Catastrophology, 35(4): 151-156
  35. 35.
    Yang C. 2013. Monitoring Regional Forest Disturbance by Remote Sensing: A Case Study of Wuning County. Nanjing: Nanjing University of Information Science and Technology
  36. 36.
    Yuan Y, Lin L, Huo L Z, Kong Y L, Zhou Z G, Wu B and Jia Y. 2020. Using an attention-based LSTM encoder-decoder network for near real-time disturbance detection. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13: 1819-1832
  37. 37.
    Zhang X Y, Friedl M A, Schaaf C B, Strahler A H, Hodges J C F, Gao F, Reed B C and Huete A. 2003. Monitoring vegetation phenology using MODIS. Remote Sensing of Environment, 84(3): 471-475
  38. 38.
    Zhang Z M, Tang D, He G J, Long T F and Wei M Y. 2020. Global 30-meter resolution burned area products over the globe based on Landsat 8 images. China Scientific Data, 5(4): 91-96
  39. 39.
    Zhao C J, Zhang P, Zhu J, Wu C R, Wang H M and Xu K L. 2019. Predicting tongue motion in unlabeled ultrasound videos using convolutional LSTM neural networks//2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)：5926-5930. IEEE, Piscataway, NJ.
  40. 40.
    Zhong X G, Mohamed A S and Florian M W. 2012. Testing the accuracy of query optimizers//Proceedings of the Fifth International Workshop on Testing Database Systems (DBTest '12. Association for Computing Machinery, 11:1-6
  41. 41.
    Zhu Z. 2017. Change detection using Landsat time series: a review of frequencies, preprocessing, algorithms, and applications. ISPRS Journal of Photogrammetry and Remote Sensing, 130: 370-384
  42. 42.
    Zhu Z, Fu Y C, Woodcock C E, Olofsson P, Vogelmann J E, Holden C, Wang M, Dai S and Yu Y. 2016. Including land cover change in analysis of greenness trends using all available Landsat 5, 7, and 8 images: a case study from Guangzhou, China (2000—2014). Remote Sensing of Environment, 185: 243-257
  43. 43.
    Zhu Z and Woodcock C E. 2014. Continuous change detection and classification of land cover using all available Landsat data. Remote Sensing of Environment, 144: 152-171

Lire l'article complet

The above content is generated by Large Model Translation. The translated content is for reference only. We do not assume any commercial or legal responsibilty for any consequences arising from the use of our website