Near-remote green: Red perpendicular vegetation index ground cover fraction estimation in cotton

Bablu Sharma, Glen L. Ritchie, Nithya Rajan

Research output: Contribution to journalArticle

9 Scopus citations

Abstract

Ground cover fraction (GCF) can be used as a proxy for leaf area index, plant radiation capture, and plant canopy characteristics in cotton (Gossypium hirsutum L.). One method of imagery-based GCF estimation is to separate plant pixels from soil pixels based on intensity of reflected green and red radiation. However, this method can be time-consuming, may be subject to bias, and is limited by image resolution. We examine a simple, image-based measure of GCF that provides a rapid measurement of crop growth based on the concept of the perpendicular vegetation index (PVI) but using two visible camera channels. The method is based on two linear relationships: one of which measures the relationship between intensity of green and red reflectance for all soil brightness values (the soil line) and another that measures green and red for 100% canopy cover. The GCF in an image is then calculated based on the mean reflectance of the image and the ratio of the image green values to that of 100% GCF from a defined soil line (GCFPVI-Green). A strong linear relationship was found between the PVI-Green and amethod of separating soil pixels from plant pixels (GCFPixelCount).The GCFPVI-Green was relatively insensitive to multiple cultivars and irrigation levels. The high correlation between GCFPVI-Green and GCFPixelCount, as well as the similarities of results between this method and previous methods based on near-infrared (NIR) and red pixel values, suggest that PVI-Green may be useful as a more timely alternative method to estimate GCF in agricultural fields.

Original languageEnglish
Pages (from-to)2252-2261
Number of pages10
JournalCrop Science
Volume55
Issue number5
DOIs
StatePublished - Sep 1 2015

Fingerprint Dive into the research topics of 'Near-remote green: Red perpendicular vegetation index ground cover fraction estimation in cotton'. Together they form a unique fingerprint.

  • Cite this