Multispectral UAV Imagery based on Normalised Difference Land-vegetation Index in image processing for Internet of Things

Abstract

In this paper, a normalised difference land-vegetation index (NDLI) from unmanned aerial vehicles (UAVs) with wireless sensor networks is demonstrated significant potential for precision agriculture. Data from a UAV equipped with a multispectral mica sense red edge camera is used as ground truth in this investigation to calibrate Sentinel imagery. By distinguishing no-green plant pixels, UAV-based NDLI enabled crop assessment at (1187x707) image pixel resolution. The reflectance value and NDVI of crops at various stages is calculated using both UAV and Sentinel-2 pictures. In this investigation UAV multispectral mapping technology gave advanced information about the physical characteristics of the studied area and better land feature delineation. According to the results, UAV data produced more accurate reflectance values than Sentinel-2 photography. The accuracy of the vegetation index, on the other hand, is not entirely dependent on the precision of the reflectance. Introduced with a simple Green Measurement Matrix (DGM) compared to NDLI produced from Sentinel-2 images, UAV-derived NDLI shows comparatively low sensitivity to vegetation coverage and is unaffected by environmental conditions. Further, it is used for the development of IoT networks by the means of UAV connected with internet for instant field study.