Camera-based eye blinks pattern detection for intelligent mouse

摘要: Human–computer interface systems provide an alternative input modality to allow people with severe disabilities access computer systems. One of the inexpensive and unobtrusive methods for this purpose is image-based eye blinks detection. Currently, available human–computer are often intrusive, limit in head rotation, require special hardware, have lighting or manual initialization. This paper presented a new robust method real-time enables interaction using “blink patterns,” which sequences long short interpreted as semiotic messages. The precise location determined automatically through multi-cues, accompanied by integration variance feature Gaussian Mixture Model classifier. detected window converted into binary image. eyelid’s distance extracted applying projection derivative function. By following finite-state machine, blink patterns can be detected. performance algorithm evaluated several frame streams. experimental results show pattern detection system real environments.