Sturdy on Non-Destructive Measurement System using Hyperspectral Imaging for Strawberry and other fruits and vegetables.

近年来，消费者对农产品质量和安全性的要求越来越高。对食品质量进行严格检测的非破坏性测量技术变得越来越重要。因此，需要更准确的质量测量技术来分选水果和蔬菜，如草莓在收获后的处理。本研究的主要目的是利用目前美国最热门的高光谱成像技术，开发一套综合的水果品质检测系统。利用高光谱成像技术对草莓、芒果、桃和茄子进行了外部和内部食品质量检测（成熟度、虫害、硬度、瘀伤、可溶性固形物含量、色素功能等）。主要研究结果如下：1.草莓、桃、芒果的品质（1）草莓在近红外（600- 1000 nm波长）上的特征光谱图像为825 nm和980 nm波长，桃在近红外（810 nm波长）上的特征光谱图像为825 nm和980 nm波长。但 ...更多信息 在芒果上没有特殊的光谱图像。（2）MR特殊频谱图像（600- 1000 nm波长）对可溶性固形物含量和成熟度的预测结果是，对于50%成熟度的草莓，三波长预测模型的相关性为0.78以上，而对于芒果和桃，五波长预测模型的相关性为0.79以上，结果表明，高光谱成像技术对水果和果蔬的外观品质和内在品质的无损评价具有重要的应用价值。紫肉甘薯和茄子色素的功能（花色苷色素的估算）紫肉甘薯花色苷色素的最强吸收波长为525 nm，茄子花色苷色素的最强吸收波长为716 nm。光谱图像的吸光度与测定的花色苷色素之间的相关系数在0.75- 0.97之间。质量综合检测系统的研制开发了一套基于高光谱图像的质量综合检测系统。这是一种简化的非破坏性类型，成本低廉。少

Recently, consumers are demanding for higher quality and better safety of agricultural produce. A non-destructive measurement technique for inspecting food quality strictly is becoming more important. Thus, more accurate quality measurement techniques are needed for sorting fruits and vegetables such strawberry during post-harvest processing. The main goal of this research is to develop a synthesis system of quality measurement for inspecting fruits using a hyperspectral imaging which is hotter technique at USA. Measurement using hyperspectral imaging was carried out an external and internal food quality inspection (ripeness, insect damage, firmness, bruises, soluble solids content, function of pigment etc. for strawberry, mango, peach and eggplant. An outline of research results was as follows.1. Quality of strawberry, peach and mango(1)The special spectral image at NIR (600-1000nm wavelength)for detecting was 825nm and 980nm wavelength on strawberry,810nm wavelength on peach. But the … More re was not special spectral image on mango.(2)The special spectral image at MR (600-1000nm wavelength)for an estimation of a soluble solids content and a ripeness was that the three-wavelength prediction model for strawberry of 50% ripe over was correlation at 0.78 over, and the five-wavelength prediction model for mango and peach was correlation at 0.79, 0.80 respectively.Hyperspectral imaging was found very useful for the non-destructive estimation of external and internal quality of fruit and fruit-vegetables.2. Function of pigments of purple-fleshed sweet potato and eggplant (an estimation of anthocyanin pigments)The most strong absorbance spectral wavelength of anthocyanin pigments was 525nm on purple-fleshed sweet potato, and 716nm on eggplant. The relationship between absorbance of spectral image and measured anthocyanin pigment was a correlation coefficient at 0.75-0.97.3. Develop of a synthesis system of quality measurementA synthesis system of quality measurement using a hyperspectral image was developed. It was a simplified non-destructive type and low cost. Less