Smartphone-based food category and nutrition quantity recognition in food image with deep learning algorithm

According to the similar nutritional properties, foods could be classified in six groups (Vegetables, Fruits, Dairy, Oils, Grains and Protein foods) and nourish human body respectively. However, people could not understand the nutrients of foods which they obtained generally. Hence, this paper propo...

Full description

Saved in:
Bibliographic Details
Main Authors: Chiun-Li Chin, Chen-Cheng Huang, Bing-Jhang Lin, Guei-Ru Wu, Tzu-Chieh Weng, Ho-Feng Chen
Format: Conference Proceeding
Language:English
Subjects:
Biomedical imaging
Image color analysis
Image recognition
Machine learning
Network architecture
Smart phones
Training
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:According to the similar nutritional properties, foods could be classified in six groups (Vegetables, Fruits, Dairy, Oils, Grains and Protein foods) and nourish human body respectively. However, people could not understand the nutrients of foods which they obtained generally. Hence, this paper proposes a system based on deep learning for training. Users take pictures on diets by their smartphones and the system will recognize both what kinds of group and how much of nutrients they will take in. With our system, users could recognize the nutrients in their diet and they can administer their health effectively. During training, we not only confirm the architecture of CNN, but also find out that the color feature of foods in the images has significant effect on the identification result about up to seventy percent of the resolution ratio.
ISSN:2377-5831
DOI:10.1109/iFUZZY.2016.8004962