Rice growing stage monitoring in small-scale region using ExG vegetation index

In this paper, a software program is developed to monitor rice growing stages. Images are required as input data for the software. Using field server equipment, the images are obtained from two rice fields located in Suphanburi and Roi Et provinces, Thailand. Each daily image covers approximately 100 × 100 m2 recorded by 720 × 480 pixels. Typically, a rice growing cycle is separated by three stages; seedling, tillering and heading. To define each stage, vegetation index is used for monitoring and analysing. In the prototype software, the vegetation index is computed from visible RGB channels. Our proposed diagram is described by three steps. a) Rice field segmentation is an initial step used to segment rice field region from the other regions (landscape, sky). b) Vegetation index computation is a measurement, which measures the levels of live green plants on the rice field region. c) Graph analysis is an algorithm used to determine and separate the rice growing stages. The experiments compared three vegetation indices; ExG-Excessive Green, NGRDI-Normalized Green Red Difference Index and ExGR-difference of ExG and ExR (Excessive Red). Relying on the images obtained from the field server, we found that the rice growing stages are able to monitor by using ExG index which is more efficient than the other two.