Share
Export Citation
Estimation of Cholesterol for Multiple Food Dish on One Plate Using Computer Vision
Hernawati
Proceedings 2023 3rd International Conference on Electronic and Electrical Engineering and Intelligent System Responsible Technology for Sustainable Humanity Ice3is 2023
Abstract
High cholesterol is a dangerous condition that can lead to heart attacks, strokes, and even death. One of the causes is consuming foods that are high in cholesterol. Therefore, it is essential for us to know the cholesterol content of every food we consume daily. In an effort to help consumers manage their food intake, we conducted research to create an Android-based application that utilizes computer vision technology. In this research, we employed a deep learning approach to estimate the cholesterol content of various types of food served on a plate. We utilized the Single Shot Multibox Detector (SSD) method with the Mobilenet v2 architecture to detect food objects in images. Subsequently, the detected images were labeled with the respective cholesterol amounts for each food object, presented as bounding boxes. To implement this application in real-time on mobile devices, we used the Tensorflow object detection framework during the extraction process. Through our object detection testing, we achieved an accuracy rate of 93%. This result demonstrates that the Computer Vision approach using the SSD Mobilenet V2 architecture is capable of effectively detecting food objects.