Multi-source remote sensing of banana LAI across full growth cycle using canopy spectral-morphological features and temporal Transformer-LSTM

Shaoming Lin

doi:10.1016/j.compag.2026.111870

Multi-source remote sensing of banana LAI across full growth cycle using canopy spectral-morphological features and temporal Transformer-LSTM

2026-05-10

0

PRE

AI

S

Shaoming Lin

C

Chao Ma

J

Juan Wang *

S

Shuchang Wang *

Y

Yuhang Zhang

Z

Zhen Gao

T

Tianpeng Ji

DOI：10.1016/j.compag.2026.111870

Original

Original request for help

Save

Abstract

En 中文

• A novel Regional Volume improves LAI estimation by excluding the under-canopy space. • The Transformer-LSTM model captures full-cycle banana LAI dynamics across growth stages. • Multi-source feature combinations achieves superior LAI estimation accuracy over single-feature. • The proposed framework maintains robust LAI estimation performance across all growth stages.

Keywords:

Banana LAI

Remote sensing

Transformer-LSTM

Canopy features

Growth cycle

Multi-source remote sensing of banana LAI across full growth cycle using canopy spectral-morphological features and temporal Transformer-LSTM

Abstract

Journal

Organization

Cited Papers

Citing Papers