AERIAL MULTISPECTRAL IMAGING FOR DETECTION AND QUANTIFICATION OF WEEDS.

Multispectral remote sensing is a new effective technique to evaluate crops phenology, plant health and the weeds presence. The study aims to exploit the possibilities of multispectral imaging for detection the weeds in winter wheat. The phenological changes of different weeds was also investigated, employing the calculation of six vegetation indexes during a period of three months. The vegetation indexes were as follows: CIG, GRNDVI, GRVI, CVI, NDWI2, NDVI. The influence of weed species on the value of the corresponding vegetation index within each reporting period was investigated. The control value was the one calculated for areas sown with wheat without weeds. It is established that the factors: phenophase and weed type have a statistically significant influence on the researched indexes. The GRVI is the only index on which plant phenophase does not have a significant effect. Regarding the NDVI values, no permanent trend for the presence/absence of a certain weed was found. In almost all phenophases, high values of the index were calculated in both weed varieties and weed free wheat. [ABSTRACT FROM AUTHOR]