The Application of Multispectral Unmanned Aerial Vehicles (UAVs) for Evaluating the Yield of Vegetable Soybean at the Chiang Mai Agricultural Research Center

Abstract

The use of multispectral imagery from a small Unmanned Aerial Vehicle (UAVs) to monitor the growth and yield of vegetable soybean had as its objective to monitor the various stages of growth associated with yield by collecting aerial growth data at 4 stages from 7, 16, 31 and 61 days after sowing (DAS), and transforming the images into the Normalized Difference Vegetation Index (NDVI) and the visible light wavebands (red green and blue, RGB) The image data was used to establish a correlation between the growth stage and the yield of vegetable soybean. The results of the study showed that the area covered by vegetable soybean increased depending on the growth stage after seed germination from a minimum 7 days and a maximum of 61 days. The green content of the leaves was lowest after 7 days and highest after 31 days. The greenness of the leaves in the RGB image tended in the same direction as the NDVI image, with values closest to -1 at 7 days and closest to 1 at 31 days. The NDVI value at 31 DAS had the best relationship with planted area (R2=0.86) with the highest value of 0.97 and the lowest of 0.77. Total production at the 31 DAS stage was in the range of 1,200 g – 1,330 g/ 2 m² (960–1,064 kg/rai) and the NDVI value was in the range of 0.65–0.79, which tended to increase the yield of fresh soybean pods depending on the NDVI value (R2=0.80).