Identification of vegetation types and its boundaries using artificial neural networks

International Workshop "Advanced Technologies in Material Science, Mechanical and Automation Engineering – MIP: Engineering – 2019"

材料科学;机械制造;原子能学

Saltykov, M.^1 ; Yakubailik, O.^2 ; Bartsev, S.^1

Institute of Biophysics, FRC KSC SB RAS, Akademgorodok 50/50, Krasnoyarsk

660036, Russia^1

Institute of Computation Modeling, FRC KSC SB RAS, Akademgorodok 50/44, Krasnoyarsk

660036, Russia^2

关键词: Boreal forests;    Mixed forests;    Multi-spectral imagery;    Satellite images;    Spectral channels;    Trained neural networks;    Vegetation index;    Vegetation type;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/537/6/062001/pdf
DOI  :  10.1088/1757-899X/537/6/062001

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

The applicability of artificial neural networks (ANN) for the identification of vegetation types using satellite multispectral imagery was studied. The study was focused on the three main vegetation types found in the south of the Krasnoyarsk Region: mixed forest, boreal forest and grassland. Sentinel-2 satellite images were used as a data source for the neural networks. It was shown that vegetation type can be identified pixel-by-pixel using 12 spectral channels and simple feed forward ANN with good quality and reliability. Analysis of the input layer of the trained neural networks allowed several spectral bands to be selected that were the most valuable for the ANN decision and not used in the classic NDVI vegetation index.