Histogram-Based Food Image Classification and Nutrition Detection Using Support Vector Machine

Abstract

Food image classification and nutrition detection have become important research areas in the field of computer vision and machine learning. With the increasing use of smart phones and digital health applications, recognizing food items from images and calculating their nutritional value can help users monitor their diet and promote healthy lifestyles. This paper presents a method for food image classification and nutrition detection using a Support Vector Machine (SVM) model. The proposed approach involves preprocessing food images, extracting visual features such as sum of the values of equalized histogram bins as color feature and classifying food types using SVM. Nutritional information, including calories, proteins, fats, and carbohydrates, is then estimated from a predefined nutrition database. Experimental results demonstrate that SVM provides high classification accuracy for various food types and can effectively support automated dietary assessment systems. Experimental evaluation demonstrates a classification accuracy of 94.6% and a nutrition estimation error below 5%, validating the reliability of the proposed approach.