Spectroscopy and Spectral Analysis, Volume. 42, Issue 6, 1939(2022)
Study on Classification and Recognition of Mountain Meadow Vegetation Based on Seasonal Characteristics of Hyperspectral Data
Figures & Tables(6)
Fig. 1. Location and typical vegetation composition in study area (Tianshan Mountains marked by black border in the map)
Fig. 2. Hyperspectral images of vegetation in different periods
(a): Green-up date; (b): Tillering stage; (c): Flowering phase; (d): Withering date
Fig. 3. Comparison of spectral curves of dominant vegetation in different seasons before and after smooth denoising
Fig. 4. Comparison of spectral curves of dominant vegetation in four key growth periods
Table 1. Modeling results of support vector machine by using hyperspectral data
View tableView in ArticleTable 1. Modeling results of support vector machine by using hyperspectral data
生育期 核函数 总体分类精度/% Kappa系数 返青期 RBF 95.88 0.95 Polynomial 96.91 0.96 分蘖(枝)期 RBF 95.19 0.94 Polynomial 97.86 0.97 花期 RBF 91.01 0.89 Polynomial 86.62 0.84 黄枯期 RBF 88.27 0.86 Polynomial 87.80 0.86
Table 2. Classification results and evaluations of multiple methods
View tableView in ArticleTable 2. Classification results and evaluations of multiple methods
生育期 分类方法 总体分类精度/% Kappa系数 返青期 BP-ANN 79.50 0.77 SAM 65.42 0.61 分蘖(枝)期 BP-ANN 91.07 0.89 SAM 77.80 0.73 花期 BP-ANN 88.87 0.87 SAM 75.39 0.72 黄枯期 BP-ANN 89.90 0.88 SAM 46.00 0.38
Tools
Get Citation
Copy Citation Text
Yi ZHENG, Yao WANG, Yan LIU. Study on Classification and Recognition of Mountain Meadow Vegetation Based on Seasonal Characteristics of Hyperspectral Data[J]. Spectroscopy and Spectral Analysis, 2022, 42(6): 1939
Paper Information
Category: Research Articles
Received: May. 12, 2021
Accepted: --
Published Online: Nov. 14, 2022
The Author Email:
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20