关键词: 高光谱成像, 成像光谱, 计算机视觉, 机器视觉, 食品, 品质检测, 图像处理, 模式识别

Abstract:

In order to meet the consumers’demands on the quality and safety of food products,it is required to apply rapid and non-invasive inspection during production and processing of food products in the modern food industry. The quality and safety issues of food products in China have raised a high attention of the whole society in recent years. By acquiring the images of food products in different spectral ranges,modern optical imaging techniques use digital image processing algorithms to extract feature information from digital images of food products and establish quantitative relationships between features and quality attributes of the products,resulting in realizing rapid,non-invasive,efficient,and low-cost assessment of food products and providing information support to automatic sorting and grading management,and to quality inspection in modern food process operations. Food Refrigeration &Computerised Food Technology ( FRCFT) in University College Dublin ( UCD) led by Professor Da-Wen Sun has carried out a series of research works on applying the advanced optical imaging techniques including computer vision and hyperspectral imaging to determine food quality in a rapid and non-invasive way. Their leading research works are well recognized by their international peers. This paper reviews the research results of UCD-FRCFT on food quality inspection using computer vision and hyperspectral imaging in the past decade.

Key words: hyperspectral imaging, imaging spectroscopy, computer vision, machine vision, food, quality inspection, image process, pattern recognition