Crispbread quality evaluation based on fusion of information from the sensor analogies to the human olfactory, auditory and tactile senses

Appreciation of food is based on the perception all five human senses; visi on, tactile, auditory, taste and olfaction. Sensory integration of chewing resistance, melting properties, crisp sound etc. will give a further percep tion of the food, which is called the mouth feel. In this report, a model e xperiment performed on crispbread is described, which was based on informat ion fusion from human sense analogies of olfaction, auditory and tactile. F ive samples of crispbread with various hardness and flour composition were placed in a special "crush chamber". While crushed, information correspondi ng to three senses could be obtained - auditory by a microphone, tactile by a force sensor and smell by leading gases from the crushed material to a g as sensor array. To evaluate the information obtained, multivariate pattern recognition methods, i.e. principal component analysis and artificial neur al nets were used to search for structure and correlation in the data. It w as shown that by using only the auditory information, the samples could not be separated. The tactile information alone could separate three samples, and the smell information one sample. By combining all sense analogues, all five samples could be separated.