Terahertz Radar and Deep Learning-Based Detection of Soft Foreign Objects in Food Products: An Automatic Inspection Approach

Abstract

This study investigates the feasibility of an automatic detection system for foreign objects in food using terahertz radar and deep learning techniques. The experimental setup comprises a terahertz radar source, beam splitters and collimators, and a terahertz receiver with a 32 x 32, 1024-pixel area scanner. Received images representing signal strength transmitted through noodles serve as input for a deep learning model after preprocessing to eliminate noise. A binary decision is then made on whether the food contains foreign objects or not. © 2023 IEEE.