Community Speakers: Janos Keresztes

    Friday, May 1, 2015 - 12:00pm - 1:00pm
    Meinel 408

    "Illumination Design and Gloss Modeling for Real-Time Food Quality Inspection Using Short-Wave Infrared Hyperspectral Imaging"


    Given an increasing global human population, a sufficient food supply is an important concern. More efficient food-quality inspection is needed in order for farmers to survive globalization at a competitive price, answering to demand while minimizing waste. In past years, researchers demonstrated hyperspectral imaging as a more reliable nondestructive quality-inspection system than traditional machine vision. More recently, the short-wave infrared spectral range enabled within hyperspectral images has provided the possibility to monitor moisture, fat, vitamins, oxidation and other biochemical content, offering new alternatives for food sorting. To deliver more robust and reliable SWIR hyperspectral imaging in industry, broadband spectra uniform illumination, high-speed and low-gloss systems are needed. Within his research at the University of Leuven (Belgium), Keresztes will describe how he designed an optimal SWIR illumination system for apples and a procedure to detect early damages in real time using SWIR hyperspectral imaging while handling glare artifacts and sensor nonuniformity. He will present his current research at the College of Optical Sciences on polarimetric and spectral bidirectional reflectance distribution function measurements to model better fruit optical properties within his illumination design.