Search

This paper describes a method for quantifying food appearance and studies the relationship between colour appearance and sensory characteristics of expected levels. Orange juice is used as an example. An experiment involving visual assessments was carried out using a calibrated digital display. The first phase of the experiment (i.e. Phase I) focused on investigating tolerance for colour as an orange juice attribute and the second phase (i.e. Phase II) concentrated on relationships between juice colours and expected sensory characteristics. Visual judgements were made of sourness, sweetness, bitterness, flavour strength and freshness. In Phase I, 174 juice colours were rendered systematically in CIELAB colour space and assessed by 15 observers. It was found that colour tolerance of orange juices can be determined using the CIELAB colour difference formula ( Δ E ab ∗ ). A colour will be accepted by the majority as natural orange juice if its colour difference against an ideal juice colour is smaller than 12.60 Δ E ab ∗ units, where the lightness, chroma and hue of the ideal orange juice colour was 67, 62 and 88°, respectively. In Phase II, observers were asked to assess the same panel of stimuli using the expected levels of the five sensory characteristics. It was found that greenish juice colours elicited greater sourness and bitterness responses. Darker juice colours were more likely to be expected to be bitter, and redder and yellower juice colours were expected to be sweeter and have stronger flavour. Fresher juices were distributed within the region of saturated yellow. These relationships were described by means of Δ E ab ∗ which reasonably explains the relationships except the cases of sourness (R2 = 0.66) and freshness (R2 = 0.66). A new colour difference formula ΔEOJ was proposed and this formula effectively improved the performance of predictions for sourness (R2 = 0.72) and freshness (R2 = 0.82). The methodology developed in this study includes a systematic study to find the “ideal” colour appearance of a particular food, application of the psychophysical method for assessing expected levels of different sensory characteristics and a method for modelling the appearance and expectation relationships. This methodology can be widely applied to optimise visually perceived expectations for other foods and products that are sensitive to visual judgements of quality.