Unmanned Aerial Vehicle-Based Multispectral Remote Sensing for Commercially Important Aromatic Crops in India for Its Efficient Monitoring and Management

Aromatic plants cultivation, processing and marketing is an upcoming agro-industry. The yields from these plants are generally governed by its good management practices of timely, suitable and precise actions against damaging factors. Remote sensing in agriculture is not a new phenomenon anymore, but using unmanned aerial vehicle (UAV, commonly known as drones) for the same is a pertinent topic these days, especially in India. Therefore, the study seeks to perform UAV-based airborne data acquisition, processing and analysis for modernised agricultural practices, finding of which may lead to generate rapid and on-demand real-time remotely sensed data for precision agriculture of commercial crops, which require more care and timely inputs as compared to conventional crops. The UAV high-resolution (1.5 cm/pixel) data were acquired from Mica Sense Altum, a 6 bands multispectral sensor, mounted over an indigenous Quad-copter (< 5 kg). With the help of processed orthoimage, the 22 plots of Rosa damascena (Damask Rose) were precisely (95% accuracy) classified into 03 categories, i.e., rose canopy, weed and open soil areas. We have also estimated digital plant count, plant height derived from canopy height model (CHM), canopy temperature and the topographic conditions of the crop plots. The digital plant counting for R. damascena planted in 4323 m(2) area took 1.2 h as compared to manual 5.94 h counting. Average plant height values derived from CHM ranged from 23-68 cm as compared to 28-71 cm manually measured heights. Results were compared with ground sampling data, with which high correlation was found in digital plant count (R-2 = 0.99) and plant height (96.69% accuracy). The derived average moderate slopes and northeast aspect suggested suitable topographic conditions required for R. damascena cultivation. The image-derived canopy temperature was compared to the relative ground-based measurements, obtaining accuracy percent of 98.54%. The outcomes are encouraging and have potential to be applied for future UAV grounded applications by farmhands.