Paddy Growing Stages Model Based on Vegetation Indices Using Ultra-High Spatial Resolution Images

Abstract

Paddy is one of the most important food sources in Indonesia. It is evidenced by the increasing number of national rice consumption averagely at 6.29% per year, particularly in 2011–2015. However, the production seems does not equally match the rise in consumption. Estimates in rice production are relatively unreliable. It is due to the uneven planting time in several areas and a conventional method applied to estimate the production. This study proposes alternative methods to estimate rice production. This study aims to analyze the paddy growing stages and determine the most optimal model to estimate the paddy growing stages based on the vegetation indices. This study used the excellence of remote sensing technology especially for paddy field monitoring, emphasizing on paddy growing stages assessment. An airborne remote sensing platform, specifically the Unmanned Aerial Vehicle (UAV) is used to map the rice field in Bekasi Regency, West Java Province. Through mapping at low altitude, the UAV can produce images with ultra-high resolution, so it is very well used for mapping the paddy growing stages with diverse characteristics. Several vegetation indices, derived from Red, Green, and Blue (RGB) bands, namely Normalized Green Red Difference Index (NGRDI), Excess Green Vegetation Index (ExG), and Visible Atmospherically Resistant Index (VARI). Furthermore, the regression model is used to obtain the most optimal model of the three vegetation indices used for estimating the paddy growing stages. The result showed that the UAV with RGB bands could be used as a sensor to determine the relationship between vegetation indices to the paddy growing stages and the most optimal model for estimating the paddy growing stages based on the vegetation indices is ExG (R2 = 0.88).