Resolution analysis for estimating human numbers by thermal array sensor system
Yusuke Taniguchi, Hiroshi Nakajima, Naoki Tsuchiya, Junichi Tanaka, Fumiji Aita, Yutaka Hata · 2014
This paper describes an estimating system of human numbers for resolution change by a thermal array sensor. In this system, the sensor is attached to ceiling of a room and acquires temperature distributions by 16 × 16 elements. The temperature distributions represent a state of temperature in the room, and they are analyzed to estimate human numbers. Temperature distributions consist of 16 × 16 pixels. The, temperature distributions are squared of each pixel temperature and reduced their resolution to 8 × 8 and 4 × 4. Then, human numbers are estimated by fuzzy inference. In our experiment, we measured temperature distribution in a room to evaluate of our system. From the results, the proposed method obtained higher accuracy than previous work in 8 × 8 and 4 × 4 resolution. Estimation accuracy of 8 × 8 resolution is better than 4 × 4 resolution.