Classification Of Nutrition Deficiency In Rice Plant Using CNN

Abstract

Nutrient deficiency often occurs in rice plants, thus affecting the level of production and quality of rice. Nutrient deficiency, in general, can be seen from the color and shape of sick leaves; therefore, it can be detected early to reduce the symptoms of nutritional deficiency in rice plants. This study classifies the symptoms of nutritional deficiency in rice plants using the Convolutional Neural Network (CNN) with ResNet 50 and ResNet 152 architectures. There are 1156 images with datasets sourced from Kaggle, divided into nitrogen (N) deficiency and Phosphorus(P) deficiency. And Potassium (K) deficiency. The dataset augmentation process used oversampling techniques to balance the data. The best results were obtained from the ResNet 50 architecture with accuracy and validation values yielding 98% and testing values 97%