Gas and odor sensors have been used in various fields, for example detection of environmental pollution, safety and quality controls in the food industry and identification of various human behaviors in a residence. Many kinds of odors occur according to human behavior. The analysis of odor characteristics has been carried out. The main odors which occur in a residence are as follows: namely odors of meal preparation, excretion, perfume and body odors. Life patterns of an elderly person in solitude can be grasped roughly by analyzing the odor sensor characteristics. Taking meals and excretion are important vital actions to human beings. The actions can be known by identifying the odors.
This study attempts to identify a meal-degree, which means the kinds of meals taken, for example a light meal, a fried meal or a meal cooked by oneself. Various kinds of odor molecules occur at every meal and the sensor patterns vary. Meal-degree can be, therefore, understood by odor sensor patterns and it is also able to give the aged some suggestions about his health care according to long-term characteristics of plural odor sensors. As for the results, the number of responded sensors are few in a light meal and the sensor-grade is also small. In contrast with this matter, all of the installed sensors respond to fried or cooked meals, and the response grade is higher than a light meal.
View full abstract