Classification Of Rice Leaf Color Into Leaf Color Chart Using LAB Color Space

Khoirul Umam,Eko Heri Susanto
DOI: https://doi.org/10.33050/ccit.v13i2.1008
2020-08-27
CCIT Journal
Abstract:Leaf Color Chart (LCC) is a measurement tool that can be used to measure the color intensity of rice leaves. The function of these measurements is to find out how many doses of fertilizer are needed by rice plants. However, readings made by human vision have a high level of subjectivity and risk of error. Therefore we need a method that can minimize errors and the level of subjectivity. One method that can be done is to classify the green color of rice leaves using LAB color space. Rice leaf image taken using a smartphone device is then extracted in RGB format. The color is then converted to LAB color space and then compared to the standard green color in the LCC. The comparison results are then used to classify the colors. The testing results show that the method has the value of accuracy, average precision, and average recall of 54.74%, 54.44%, and 51.16% respectively. Therefore the method can only classify correctly half of the data testing.
What problem does this paper attempt to address?