Abstract
The use of a room has been visualized for a series of seven days based on the measurement of sound. Although sneaking in and out of a room was difficult to measure with an index of the sound pressure, daily activities including such a sneaking action could be detected by acquiring the rapid change of the higher harmonic components of sound, superimposed on the slow-varying noise under the acoustic circumstances. Because the higher spectral components are associated with the number of the maximum and minimum of the filtered data, the extremal sampler developed in the previous study has been utilized for monitoring daily noise. The number of extrema is dependent of the acoustic environment, so that the size of the extremal files varied from less than 30% on weekdays to less than 20% over the weekend, as compared with that of conventional recording. The author suggests that the proposed recording at a bandwidth of 80-5, 120 Hz could search for a sign of activities in a large amount of the worthless noise for home security use.
