DUPLICATE REDUCTION FOR FACE DETECTION

摘要

In one embodiment, a method for detecting faces in video image frames is implemented on a computing device including: comparing current image frames to previously processed image frames to determine similarity, if a current image frame and a previously processed image frame are dissimilar, comparing a location within the current image frame for at least one detected facial image to a location within an associated image frame for at least one most recently stored facial image stored in a most recently used (MRU) cache, if the compared locations are dissimilar, comparing the at least one detected facial image to the at least one most recently stored facial image stored in the MRU cache to determine similarity, and storing the at least one detected facial image in the MRU cache if the at least one detected facial image and the at least one most recently stored facial image are not dissimilar.

主权项

1. A method for detecting faces in video image frames, the method implemented on a computing device and comprising:
comparing a current image frame to a previously processed image frame to determine similarity based on a threshold percent of deviance between said current and previously processed image frames; discarding said current image frame if said current image frame and said previously processed image frame are similar according to said comparing of image frames; if said current image frame and said previously processed image frame are not similar according to said comparing of image frames, detecting at least one detected facial image in said current image frame; comparing a location within said current image frame for said at least one detected facial image to a location within an associated image frame for at least one most recently stored facial image stored in a most recently used (MRU) cache; discarding said at least one detected facial image if said location with said current frame and said location within an associated image frame are similar according to said comparing of locations; if said location within said current image frame for said at least one detected facial image and said location within an associated image frame are not similar according to said comparing of locations, comparing said at least one detected facial image to said at least one most recently stored facial image stored in said MRU cache to determine similarity based on a threshold percent of deviance between said detected and most recently stored facial images; discarding said at least one detected facial image if said at least one detected facial image and said at least one most recently stored facial image are similar according to said comparing of facial images; and storing said at least one detected facial image in said MRU cache if said at least one detected facial image and said at least one most recently stored facial image are not similar according to said comparing of image frames.