Crop Discrimination Using Multispectral Sensor Onboard Unmanned Aerial Vehicle

Abstract

This paper presents the results and observations made from pilot exercises on crop discrimination using Parraot Sequoia multispectral sensor onboard a Unmanned Arial Vehicle (UAV). The exercise was carried out in two selected sites in West Jaintia Hills district of Meghalaya state with mixed horticultural crops. Parraot Sequoia sensor having four bands has been found to be effective in crop discrimination based on variation in spectral response and crop height. Discrimination of three horticultural crops viz. banana, orange, and plum and the neighbouring bamboo grooves were evaluated using three commonly used indices viz., Normalized Difference Vegetation Index (NDVI), Normalized Difference Red Edge Index (NDRE) and Green Normalized Difference Vegetation Index (GNDVI). NDVI and GNDVI showed nearly similar spectral response, whereas separability among the crops marginally improved with the use of NDRE. The percent variations of spectral response for orange and bamboo were 14 and 19 in terms of NDVI and GNDVI respectively, whereas the same value is 49 in case of NDRE. Similarly percent variations of spectral response for banana and bamboo were 7 and 15 in terms of NDVI and GNDVI against 27 in case of NDRE. In the second study site, variation in Digital Surface Model (DSM) and Digital Terrain Model (DTM) was generated to discriminate different crops grown at different elevation. The difference in the surface model has helped in discriminating different crops and other neighboring vegetation with rice recording difference of − 0.05 to 0.05, followed by maize (0.05–0.5) and pineapple (0.5–10).