The classification of chili (Capsicum annuum L.) powder quality by using image processing and artificial neural networks

The purpose of this study was to determine the relationship between the quality of chili powder with the color elements of the image and develop Artificial Neural Networks (ANN) architecture for the chili powder classification process. The chili (Capsicum annuum L.) powder samples were divided into three groups, namely 90 samples for training, 30 samples for validation, and 15 samples for testing. The images of chili powder were captured by using a webcam camera. Subsequently, the images were processed by using digital image processing to obtain the color and texture features for ANN input. The results showed that the elements of image colors used in the classification of chili powder quality were a, green, red, and hue had a very strong relationship. The ANN architecture used had three layers, namely the input layer comprised of 4 neurons (a, green, red, and hue), the hidden layer comprised of 8 neurons, and the output layer comprised of 2 neurons in the form of chili powder quality class with an accuracy of 93.33 %.