Classification and Assessment of the State of Mixed Forests from Very High Spatial Resolution Airborne Images

Complete text of the article:

Download article (pdf, 2.8MB )

UDС

528.854 (630.4)

DOI:

10.17238/issn0536-1036.2019.5.9

Abstract

At present, the invasion by Ussuri polygraphus (Polygraphus proximus Blandf) is considered as one of the main factors of large-scale drying of Siberian forests. The appearance of this new organism in fir trees has led to seriously worsening their condition and a variety of ecological effects in taiga ecosystems. The strong decrease of natural biological diversity, forest productivity, changes in the composition and structure of tree and subordinate layers may occur in the centers of mass reproduction. In this paper, we propose a method for determination of category of forest damage from very high spatial resolution color airborne images (5–10 cm per pixel) using machine learning methods. The method includes the stages of preprocessing, segmentation of crowns of individual trees, the classification and assessment of the forest damage in accordance with conventional standards. The images of several test plots of Stolby Nature Reserve (Krasnoyarsk Territory), obtained with the help of equipment installed on unmanned aerial vehicles DJI Phantom 3 Pro and Yuneec Typhoon H in May 2016, were used for testing the method proposed. The filtering method proposed for the stage of constructing a training set made it possible to increase the accuracy at the classification stage. The substantiation of division of the three main classes of objects into subclasses using cluster analysis is given. The presence of subclasses is caused by presence of various tree species in the test plot. A comparison of the efficiency of various supervised classification methods used for solving this problem is performed. It is shown that all the considered methods allow us to achieve a sufficiently high accuracy, about 95%. The calculation of the Cohen’s kappa coefficient shows that the classifications carried out with the help of all the considered methods have excellent agreement with the expert data. The analysis of the stability of training is carried out. Estimates of the total probability of error obtained by methods of cross-validation and resubstitution differ by less than 0.1%, which indicates the absence of the problem of overtraining. The joint analysis of accuracy and processing speed has shown that it is most appropriate to use the normal Bayesian classifier. High classification accuracy allows us to obtain estimates of 6 categories of forest damage in the test plot. The results obtained can be potentially used by regional forest management services.

Authors

E.V. Dmitriev^1,2, Candidate of Physics and Mathematics, Senior Research Scientist;  ResearcherID: E-4794-2014, ORCID: 0000-0001-5363-3934
V.A. Kozub², External Student
P.G. Melnik^3,4, Candidate of Agriculture, Assoc. Prof.; ResearcherID: E-7644-2014
A.A. Sokolov⁵, Candidate of Physics and Mathematics; ORCID:0000-0001-9236-5864
A.N. Safonova^6,7, PhD Student; ResearcherID: F-6811-2019, ORCID: 0000-0002-3290-2717

Affiliation

¹Marchuk Institute of Numerical Mathematics of the Russian Academy of Sciences, ul. Gubkina, 8, Moscow, 119333, Russian Federation; e-mail: yegor@mail.ru
²Moscow Institute of Physics and Technology (National Research University), per. Institutskiy, 9, Dolgoprudny, Moscow Region, 141701, Russian Federation; e-mail: postbox-kozub@yandex.ru
³Mytishchi Branch of Bauman Moscow State Technical University, ul. 1-ya Institutskaya, 1, Mytishchi-5, Moscow Region, 141005, Russian Federation; e-mail: melnik_petr@bk.ru
⁴Institute of Fоrеst Science, Russian Academy of Sciences, ul. Sovetskaya, 21, Uspenskoe, Moscow Region, 143030, Russian Federation; e-mail: melnik_petr@bk.ru
⁵Laboratoire de Physico-Chimie de l’Atmosphère Université du Littoral Côte d’Opale, Maison de la Recherche en Environnement Industriel 2, 189A, Avenue Maurice Schumann, 59140 Dunkerque, France; e-mail: anton.sokolov@univ-littoral.fr
⁶Siberian Federal University, prosp. Svobodny, 79, Krasnoyarsk, 660041, Russian Federation; e-mail: safonova.nastya1@gmail.com
⁷Soft Computing and Intelligent Information Systems research group, University of Granada, E-18071 Granada, Spain; e-mail: safonova.nastya1@gmail.com

Keywords

remote sensing, very high resolution images, forests, forest state assessment, forest degradation, stem pests, pattern recognition, thematic image processing

For citation

Dmitriev E.V., Kozub V.A., Melnik P.G., Sokolov A.A., Safonova A.N. Classification and Assessment of the State of Mixed Forests from Very High Spatial Resolution Airborne Images. Lesnoy Zhurnal [Forestry Journal], 2019, no. 5, pp. 9–24. DOI: 10.17238/issn0536-1036.2019.5.9

References

1. Akulov E.N., Kulinich O.A., Ponomarev V.L. Poly-graphus proximus is a New Invasive Pest of Softwood Forests in Russia. Zashchita i karantin rasteniy, 2011, no. 7, pp. 34–36.

2. Gninenko Yu.I., Klyukin M.S. Polygraphus proximus in the Territory of Russia. Zashchita i karantin rasteniy, 2011, no. 11, pp. 32–34.

3. Kerchev I.A. Ecology of Four Eyed Fir Bark Beetle Polygraphus proximus Blandford (Coleoptera; Curculionidae, Scolytinae) in the West-Siberian Region of Invasion. Rossiyskiy zhurnal biologicheskikh invaziy [Russian Journal of Biological Invasions], 2014, no. 2, pp. 80–94.

4. Kobel’kov M.E., Chukanov M.A., Khotin D.V. Status Categories of the Main Forest Forming Species of the Moscow Region. Мoscow, 2000. 40 p.

5. Krivets S.A., Kerchev I.A., Bisirova E.M., Pashenova N.V., Demidko D.A., Pet’ko V.M., Baranchikov Yu.N. Four-Eyed Fir Bark Beetle in Siberian Forests (Distribution, Biology, Ecology, Detection and Survey of Damaged Stands): Study Guide. Tomsk, 2015. 48 p.

6. Krivolutskaya G.O. Family Scolytidae, Bark Beetles. Key to the Insects of the Russian Far East. Vol. 3, part 3. Vladivostok, Dal’nauka Publ., 1996, pp. 312–317.

7. Mandel’shtam M.Yu., Popovichev B.G. Annotated List of Bark Beetles (Coleoptera, Scolytidae) of Leningrad Province. Entomologicheskoye obozreniye [Entomological Review], 2000, vol. 79, iss. 3, pp. 599–618.

8. Chilakhsaeva E.A. First Record of Polygraphus proximus (Coleoptera: Scolytidae) in Moscow Province. Byulleten’ Moskovskogo obshchestva ispytateley prirody. Otdel biologicheskiy [Bulletin of Moscow Society of Naturalists. Biological Series], 2008, vol. 113, iss. 6, pp. 39–42.

9. Barbakh W., Fyfe C. Online clustering algorithms. International Journal of Neural Systems, 2008, vol. 18, no. 3, pp. 1–10.

10. Dietterich T.G., Bakiri G. Solving Multiclass Learning Problems via Error-Correcting Output Codes. Journal of Artificial Intelligence Research, 1995, vol. 2, pp. 263–286.

11. Duda R.O., Hart P.E., Stork D.G. Pattern Classification. 2nd Edition. New York, NY, Wiley-Interscience, 2000. 688 p.

12. Freidman J.H., Bentley J.L., Finkel R.A. An Algorithm for Finding Best Matches in Logarithmic Expected Time. ACM Transactions on Mathematical Software, 1977, vol. 3, iss. 3, pp. 209–226. DOI: 10.1145/355744.355745

13. Hastie T., Tibshirani R., Friedman J. The Elements of Statistical Learning. New York, NY, Springer, 2001. 536 p. DOI: 10.1007/978-0-387-21606-5

14. Jing L., Hua B., Noland T., Li J. An Individual Tree Crown Delineation Method Based on Multi-Scale Segmentation of Imagery. ISPRS Journal of Photogrammetry and Remote Sensing, 2018, vol. 70, pp. 88–98.

15. MacQueen J.B. Some Methods for Classification and Analysis of Multivariate Observations. Proceedings of the 5th Berkeley Symposium on Mathematical Statistics and Probability. Vol. 1: Statistics. Berkeley, CA, University of California Press, 1967, pp. 281–297.

16. Niijima Y. Revision und Neubeschreibung der Polygraphus-Arten (Coleoptera, Ipidae) in Japan. Insecta Matsumurana, 1941, vol. 15(4), pp. 123–135.

17. Nobuchi A. Bark-Beetles Injurious to Pine in Japan. Bulletin of the Government Forest Experiment Station, 1966, vol. 185, pp. 1–49.

18. Safonova A., Tabik S., Alcaraz-Segura D., Rubtsov A., Maglinets Y., Herrera F. Detection of Fir Trees (Abies sibirica) Damaged by the Bark Beetle in Unmanned Aerial Vehicle Images with Deep Learning. Remote Sensing, 2019, vol. 11(6), art. 643, pp. 1–19. DOI: 10.3390/rs11060643

Received on July 30, 2019