Limited-view Prototype Design for Radar-based Fruit Imaging

The need for non-destructive food testing has received increasing attention with the rapid global growth in food demand and consumption over the past decades. Current limitations in assessing fruits’ internal quality motivates the requirements of enhancing existing quality assessment procedures in order to decrease the product wastage and increase the safety of the consumers. Microwave imaging is an emerging non-ionizing and non-invasive technology for various applications. This paper investigates and presents the effects of reducing the number of views on the capability of a Huygens based algorithm to detect and locate the seeds inside the fruits. The results show that limiting the number of receiving views to 180 degrees does not affect the resolution of the detected target while further reduction of the views to 90 degrees only causes a minor offset in localization. This reduction in the number of receiving positions makes way for the design of a faster, cheaper and less complex food imaging prototype. Consequently, we present the design of an automated limited-view data acquisition system tailored for food imaging applications.