Dipartimento di Scienze degli Alimenti, Univ of Udine, via Sondrio 2/A, 33100 Udine, Italy.
J Food Sci. 2012 Jul;77(7):S258-62. doi: 10.1111/j.1750-3841.2012.02704.x.
Fruit salads stored for different lengths of time as well as their images were used to estimate sensory shelf life by survival analysis. Shelf life estimates obtained using fruit salad images were longer than those achieved by analyzing the real product. This was attributed to the fact that images are 2-dimensional representations of real food, probably not comprehensive of all the visual information needed by the panelists to produce an acceptability/unacceptability judgment. Images were also subjected to image analysis and the analysis of the overall visual quality by a trained panel. These indices proved to be highly correlated to consumer rejection of the fruit salad and could be exploited for routine shelf life assessment of analogous products. To this regard, a failure criterion of 25% consumer rejection could be equivalent to a score 3 in a 5-point overall visual quality scale.
Food images can be used to assess product shelf life. In the case of fruit salads, the overall visual quality assessed by a trained panel on product images and the percentage of brown pixels in digital images can be exploited to estimate shelf life corresponding to a selected consumer rejection.
通过生存分析,使用不同储存时间的水果沙拉及其图像来估计感官货架期。使用水果沙拉图像获得的货架期估计值比通过分析实际产品获得的估计值更长。这归因于这样一个事实,即图像是真实食物的二维表示,可能无法全面涵盖品鉴员做出可接受/不可接受判断所需的所有视觉信息。图像还经过了图像分析以及经过训练的品鉴小组对整体视觉质量的分析。这些指标与消费者对水果沙拉的拒绝高度相关,可用于类似产品的常规货架期评估。在这方面,25%的消费者拒绝可以等同于 5 分制整体视觉质量量表中的 3 分。
可以使用食品图像来评估产品的货架期。在水果沙拉的情况下,可以利用经过训练的品鉴小组对产品图像的整体视觉质量进行评估,以及数字图像中棕色像素的百分比来估计与选定的消费者拒绝相对应的货架期。