摘要 |
<P>PROBLEM TO BE SOLVED: To provide a technique for discriminating the sleeping state of a bedded person with high precision. <P>SOLUTION: A feature quantity calculation part 14 calculates a plurality of feature quantities on the basis of a pulse wave. A sleeping state discrimination part 15 adapts the feature quantities to discrimination rules learned using statistical learning on the basis of teacher data to discriminate the sleeping state of the bedded person. The feature quantity calculation part 14 has a pulsation interval data calculation part 14b for calculating pulsation interval data showing the pulsation interval of the heart from the pulse wave, a pulsation frequency characteristic calculation part 14c for calculating the pulsation frequency characteristics of the heart from the pulsation interval data, a power spectrum calculation part 14d for calculating the power spectrum of the pulsation frequency characteristics, and an inclination calculation part 14e for calculating the inclinations in the high frequency regions of the power spectrum expressed on both logarithmic axis planes as the feature quantities. <P>COPYRIGHT: (C)2013,JPO&INPIT |