Archives

Hue Saturation Value (HSV) Color Detection to Identify Nutrient Deficiencies in Chili (Capsiumannuum L) using K-Nearest Neighbor


Arie Qurania , Prihastuti Harsani, Triastinurmiatiningsih, Dona Putra Por
Abstract

Along with the increasing demand for the use of chili in Indonesia with increasing population growth causes high chili prices on the market. Chili is widely used as a spice and spice for cooking because it has a distinctive taste and aroma that will add flavor to the cuisine [1]. Red chili (Capsicum annuum L.) is a horticultural plant that is widely used for food needs. Chili trees need macro elements such as Nitrogen (N), Oxygen and Hydrogen, while the micro elements needed include Potassium (K), Calcium (Ca), Magnesium (Mg). Nutrient deficiency in chili plants results in disruption of plant development such as yellowing of leaves and easy to fall, fragile chili and affect the quality of chili. Nutrient deficiencies in chili plants if not immediately addressed will result in plant death and decreased quality of chili production. This study will identify nutrient deficiencies in chilli plants using color extraction on leaves with Hue Saturation Value (HSV) color characteristics and identification using k-Nearest Neighbor (KNN). Classification of nutrient deficiencies in this study is the lack of N + P and lack of P + K in chili leaves. Leaf image data is divided into 300 training data consisting of 100 normal leaf images, 100 leaf images lacking N + P elements and 100 P + K nutrient deficient images, while for training data divided into 50 normal leaf image data, 50 lacking leaf images nutrient elements N + P and 50 nutrient deficiency P + K. Accuracy results using k = 3 and 5 resulted in an accuracy value of 82% for the identification of N + P deficient leaf image and 74% for identification of P + K nutrient deficiency.

Volume 12 | 06-Special Issue

Pages: 103-108

DOI: 10.5373/JARDCS/V12SP6/SP20201012