학술논문

Ingredient segmentation with transparency
Document Type
Conference
Source
2023 IEEE/SICE International Symposium on System Integration (SII) Symposium on System Integration (SII), 2023 IEEE/SICE International. :1-5 Jan, 2023
Subject
Bioengineering
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Transportation
Deep learning
Training data
Food industry
System integration
Quality control
Manufacturing
Language
ISSN
2474-2325
Abstract
In the food industry, the demand for the extraction of food ingredients from images is increasing for quality control and appearance evaluation of products in the manufacturing field. Therefore, recognition of food ingredients distribution using deep learning has been proposed. However, this method still has a problem of misrecognizing transparent ingredients as other ingredients. To solve this problem, we propose a method for generating training data that includes the transparency of food ingredients in order to improve the accuracy of food ingredient identification.