Image Segmentation of Healthy Food Using Hue Saturation Value

This study aims to distinguish nine types of food based on the pixel intensity of the HSV image of the food.
The segmentation method used is Hue, Saturation and Value (HSV) and compares its performance with RGB and
RGB+HSV combinations. The segmentation method used is Hue, Saturation and Value (HSV) and compares its
performance with RGB and RGB+HSV combinations. The data used in this study is image of green bean porridge,
nuggets, brown rice, white rice, kale, sausage and rendang. The segmentation begins by converting 162 food photos into
RGB images, then converting them into HSV images. The image segmentation results were then classified using the
KNN algorithm and tested at k = 3, 5 and 7. This is indicated by the value of accuracy, precision. Recall, and F1-scores were 96%, 94%, 94% and 94%, respectively.