A Robust Approach Using Fuzzy Logic for the Calories Evaluation of Fruits

The necessity for monitoring food calorie intake is becoming imperative, in order to prevent obesity and adopt healthy food habits. This work aims in aiding dieticians, physicians, and patients to measure their daily calorie intake by manually capturing multiple fruit images and by feeding them to the calorie measurement system which utilizes Adaptive Neuro- Fuzzy Inference System (ANFIS). This classifier is used for identification and classification of fruit type. The mass of acquired fruits is estimated using image processing techniques to calculate the relative calories present, according to the food portion nutrition tables. Our system displays the type of each of the fruits present in the multiple fruit dataset, as well as their corresponding calories present in it and the total calories of fruits in the multiple fruit image. The results obtained are shown to have better calorie estimation of fruits by utilizing ANFIS classifier and color histogram feature extraction techniques.