Rastreamento em vídeo das características da pupila / Video tracking of pupil features


IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia




The study of the pupil s movements has signicant clinical interest because it acts as an objective indicator of retinal sensitivity to light and, consequently, of the optic nerve. Its permanent oscillation is the result of balance between opposing ows of the sympathetic and parasympathetic nervous system. The pupil s state of an individual allows the diagnosis of various diseases, including sleep disorders (narcolepsy), schizophrenia, Alzheimer s, among others. This project presents a stable prototype device capable of tracking in video, the pupil area and detects its variation on a time scale due to the lighting in uence. The implemented tracking algorithm in a simple interface program is based on the detection method of contour, where the pupil is approximated by circles frame by frame, using the Hough transform for circles. All observed variables had the expected behavior when compared with their main theoretical models. The initial size of the pupil in relation to the iris, before the light stimulus, was between 24 and 39%, and the highest values for lighting around the red (550 - 650 nm) and green (450 - 600 nm). The same relationship for the situation of maximum constriction was between 17 and 26%. The average latency time ranged from 388 to 110 ms, while the average latency to maximum constriction ranged from 2 s to 500 ms. The speed of constriction varied from 0:75 to 3:6 mm=s while the speed of expansion ranged from 0:25 to 1:47 mm=s. Both rates reached the highest values in the intermediate light intensities referring to the lights around the red (550 - 650 nm) and green (450 - 600 nm), conrming the relationship with the maximum amplitude.


matching pupila transformada de hough vídeo tracking engenharia eletrica pupila (olhos) hough transform matching pupil video tracking

Documentos Relacionados