Image segmentation and classification with application to dietary assessment using BMI-calorie calculator
Abstract
Nowadays, people are more interested in their health by maintaining a proper diet. Today’s lifestyle causes obesity and malnutrition in humans because of an uncontrolled diet. This paper proposes the health monitoring system using the body mass index (BMI) calorie calculator, which guides people to take proper calories from their daily diet. The image processing steps segmentation, features extraction, and recognition are used in the dietary assessment to identify the food items. The improved performance of the multi-hypotheses image segmentation (MHS) and feed-forward neural network (FFNN) classifier for nutritional assessment was evaluated using macro average accuracy (MAA) and standard accuracy (SA) metrics, which provide an enhanced classification rate.