Discrimination of beef muscles by multispectral imaging
Ait-Kaddour, Abderrahmane | Jacquot, Sophie | Dubost, Annabelle | Micol, Didier | Picard, Brigitte | Listrat, Anne
The potential of multispectral VIS-NIR imaging to identified beef muscles in relation with their type and breed origin was examined in the present study. Samples (120 Longissimus thoracis, Biceps femoris and Semimembranosus muscles of three breeds of young bulls (Limousin, Blond d’Aquitaine, Aberdeen Angus) were investigated by a multispectral device. A total of 10,279 images (i.d. 541 cube images) were collected with the nineteen emitting LEDs (405 to 1050 nm). The muscle image cubes were analyzed by considering mean spectral data and image shape features from co-occurrence and difference of histogram matrices. The results of the PLSDA performed on image texture features and spectral data showed good classification depending on the muscle classes considered (i.e. muscles types and animal breed). This study demonstrated the promising potential of the VIS-NIR images to authenticate beef muscles.
Показать больше [+] Меньше [-]