Estimating Leaf Carotenoid Concentration of Ginger in Different Layers Based on Discrete Wavelet Transform Algorithm - Archive ouverte HAL Access content directly
Conference Papers Year : 2019

Estimating Leaf Carotenoid Concentration of Ginger in Different Layers Based on Discrete Wavelet Transform Algorithm

(1)
1
Qinhong Liao
  • Function : Author
  • PersonId : 1047055

Abstract

Ginger is one of the very important industrial crops in southwest, China. Accurate estimation of its leaf carotenoid concentration (LCC) is important to assess ginger photosynthetic capacity and direct the precision agriculture management. This study focused on introducing a new approach for estimating the LCC of ginger leaves in different leave layers. First, five commonly used vegetation indices (PSSR, PSND, CRI550, CRI700, BRI) were performed to estimate the LCC. The PSSR got a better result with the higher estimation accuracy (R2 = 0.46). Second, the discrete wavelet transform algorithm (DWTA) was used to extract the wavelet feature vectors for estimating the LCC. The result showed that the most sensitive wavelet feature vector was in the sixth decomposition scale. The highest estimation accuracy (R2) was 0.86 for the lower leaf layer. Compared with those vegetation indices, the estimation accuracy (R2) improved 46.5%–71.1%, which indicated that the LCC of ginger in different leave layers can be accurately estimated by DWTA.
Fichier principal
Vignette du fichier
478291_1_En_16_Chapter.pdf (647.81 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02124260 , version 1 (09-05-2019)

Licence

Attribution - CC BY 4.0

Identifiers

Cite

Qinhong Liao. Estimating Leaf Carotenoid Concentration of Ginger in Different Layers Based on Discrete Wavelet Transform Algorithm. 11th International Conference on Computer and Computing Technologies in Agriculture (CCTA), Aug 2017, Jilin, China. pp.152-158, ⟨10.1007/978-3-030-06137-1_16⟩. ⟨hal-02124260⟩
22 View
13 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More