Séminaire de Dr. Vu Hai, chercheur permanent de l'Institut MICA - Date : jeudi 2 juin 15h00 - Lieu : salle séminaire, 9ème étage, Institut MICA, Hanoi Unviersity of Science and Technology

Intervenant :
Dr Vu Hai, chercheur permanent du Département "Computer Vision" de l'Institut MICA

Date : jeudi 2 juin, 15h00
Lieu : salle "seminar room", 9ème étage, Bâtiment B1, Institut MICA, Hanoi University of Science and Technology
Langue : le séminaire sera présenté en anglais

Résumé/abstract:
The cost and size of Hyperspectral Imaging (HSI) systems have reduced significantly in recent years, and as a result this technique is becoming widely used in food industry and in food sciences applications. Through a collaborative research project between MICA (HUST) and EEE (University of Strathclyde), the HSI system has been adopted for rice seed quality inspection. The proposed system aims to automatically detect unwanted/unclean grains in large samples of rice seeds. At the first step, data acquisition consisting of camera setup, data calibration and normalization are taken carefully. The datasets of common rice seed species from Vietnam are collected with both NIR and VIS cameras. Then to examine performance of the HSI technique, various classification schemes are deployed. In particularly, spectral features at seed-based and pixel-based levels are utilized. The preliminary results have showed potential advantages of the HSI techniques comparing with physical grain properties such as size, shape color and texture using conventional RGB cameras. In this talk, a brief introduction of the Newton Fund programme is also given.