Classification of Macronutrient Deficiency in Chili Leaves using Support Vector Machine

Abstract

Chili is a horticultural crop that has high economic value in Indonesia. The productivity level of chili in the country is not proportional to the level of consumption, one of the causes is malnutrition. Each plant requires different amounts of macronutrients and micronutrients to support plant growth and development. Chili plants that lack or excess macronutrients show different visual symptoms. Digital Image Processing is a non-destructive method that is useful for determining plant health conditions based on visual symptoms of chili leaves. The combination of digital image processing and learning methods such as the Support Vector Machine (SVM) helps classify the types of macronutrient deficiencies in order to obtain a nutrient solution. In this study, there are several stages in determining macronutrient deficiencies in chili plants, namely image acquisition, pre-processing, feature extraction, to classification using SVM with several kernels. Based on the experimental results in this study, the SVM method can help modern farmers to determine the health condition of plants non-destructively with 97.76% accuracy using a Polynomial kernel. Applying this system to an intelligent farming system is expected to support the realization of precision agriculture in Indonesia.