Articles
Mapping apple canopy attributes using aerial multispectral imagery for precision crop inputs management
Article number
1346_68
Pages
537 – 546
Language
English
Abstract
Spatiotemporal mapping of tree-crop canopy-specific attributes, e.g., light interception and leaf area index (LAI), is critical for precision orchard management operations.
Conventional approaches include manual tape measurements of the canopy extents or ground based imaging/sensing systems.
But those offer characterization of only few trees in the orchard block per unit time and therefore lead to restricted sampling accuracies.
This study assessed the applicability of high resolution (7 cm pixel‑1) aerial multispectral imagery to map orchard-scale fraction of photosynthetically active radiation interception (FPARI) and LAI of apple tree canopies trained to a modern V-trellis architecture (cultivar: Buckeye Gala, rootstock: B.9, age: 12 years). Five imaging flights were conducted in the 2020 season, and four vegetation indices (normalized difference vegetation index [NDVI], green NDVI [GNDVI], soil adjusted vegetation index [SAVI], and LAI) were extracted.
Measurements of FPARI were collected as ground reference using line and quantum light sensors for 20 sample trees, aligned with all the aerial missions (Total 100 samples). LAI was calculated as another ground reference from measured FPARI. The aerial imagery derived NDVI, SAVI and LAI were found to have strong relationship with the ground-measured FPARI (r: 0.74-0.75, p<0.05) and LAI (r: 0.72-0.73, p<0.05). Results suggest that any of those spectral indices could be suitable for mapping and understand the spatiotemporal canopy distribution, vigor, and stress characteristics for precision crop inputs management.
Conventional approaches include manual tape measurements of the canopy extents or ground based imaging/sensing systems.
But those offer characterization of only few trees in the orchard block per unit time and therefore lead to restricted sampling accuracies.
This study assessed the applicability of high resolution (7 cm pixel‑1) aerial multispectral imagery to map orchard-scale fraction of photosynthetically active radiation interception (FPARI) and LAI of apple tree canopies trained to a modern V-trellis architecture (cultivar: Buckeye Gala, rootstock: B.9, age: 12 years). Five imaging flights were conducted in the 2020 season, and four vegetation indices (normalized difference vegetation index [NDVI], green NDVI [GNDVI], soil adjusted vegetation index [SAVI], and LAI) were extracted.
Measurements of FPARI were collected as ground reference using line and quantum light sensors for 20 sample trees, aligned with all the aerial missions (Total 100 samples). LAI was calculated as another ground reference from measured FPARI. The aerial imagery derived NDVI, SAVI and LAI were found to have strong relationship with the ground-measured FPARI (r: 0.74-0.75, p<0.05) and LAI (r: 0.72-0.73, p<0.05). Results suggest that any of those spectral indices could be suitable for mapping and understand the spatiotemporal canopy distribution, vigor, and stress characteristics for precision crop inputs management.
Publication
Authors
A.K. Chandel, A.P. Rathnayake, L.R. Khot
Keywords
apple canopy, light interception, leaf area index, small unmanned aerial system, vegetation indices
Online Articles (107)