Return to search

Spectral differentiation of Cannabis sativa L from maize using hyperspectral indices.

Cannabis sativa L. is a drug producing crop that is illegally cultivated

in South Africa. The South African Police Service (SAPS) use aerial

spotters on low flying fixed wing aircrafts to identify cannabis from

other land cover. Cannabis is usually intercropped with maize to

conceal it from law enforcement officers. Therefore the use of remote

sensing in identifying and monitoring cannabis when intercropped with

maize and other crops is imperative.

This study aimed to investigate the potential of hyper spectral indices to

discriminate cannabis from maize under different cropping methods,

namely, monocropped and intercropped. Cannabis and maize were

grown in a greenhouse. The spectral signatures were measured in a

dark room environment. Green pigments (chlorophyll and carotenoid)

from the treatments were also measured. These pigments were then

compared with their respective indices. Photosynthetic reflective index

(PRI) and Carotenoid Reflective Index (CRI) were two of the indices

used to discriminate cannabis from maize using carotenoid content

while the Red Edge Position (REP) and the narrow band Normalized

Difference Vegetation Index (NDVI) used chlorophyll content and

morphological differences respectively to discriminate the two plant

species.

CRI and NDVI proved to be capable of identifying cannabis under the

two cropping conditions. NDVI showed a 25% spectral over lap for the

monocropped treatments and 60% over lap for the intercropped

treatments. CRI displayed 18% and 58% over lap for the monocropped

and intercropped treatments, respectively. As a result CRI emerged as

the most suitable index for discriminating cannabis from maize. With

proper calibration of airborne or space borne imagery, the study offers

potential to detect cannabis using remote sensing technology. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2010.

Identiferoai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:ukzn/oai:http://researchspace.ukzn.ac.za:10413/9848
Date31 October 2013
CreatorsSibandze, Phila.
ContributorsMutanga, Onisimo.
Source SetsSouth African National ETD Portal
Languageen_ZA
Detected LanguageEnglish
TypeThesis

Page generated in 0.0016 seconds