Incremental support vector machine framework for visual sensor networks
EURASIP Journal on Applied Signal Processing
Dynamic classification for video stream using support vector machine
Applied Soft Computing
| Hi-index | 0.01 |
Position tracking of rodents is useful and necessary to help elucidate the behavioral and physiological effects of genetic mutations, drug action, and environmental stimuli. In this paper we describe a real-time system developed to allow continuous overhead video monitoring of rodent behavior in a home cage environment, either in a daylight condition (light-cycle) using standard visible illumination or a night-time condition (dark-cycle), using overhead near infrared illumination (NIR). Due to the lack of research on the effects of NIR on rodent behavior, we also examined open-field locomotor activity under 880 nm and 940 nm wavelengths of NIR, as well as visible white light and a "dark" condition consisting of a very dim level of NIR. The experimental result validated the described system and robustly tracked the target rodent in the light cycle, and for high contrast conditions in the dark cycle.