Simulating a fluttering shutter using video data to eliminate motion blur

摘要

A system and method 300 for simulating a fluttering shutter from video data 301, comprising using a video camera (411, Figure 4) to produce a plurality of captured video frames 301, scaling the video frames 306 according to a sequence of weights 308, combining the scaled sequence of video frames 309 to generate a composite image 310 with a coded motion blur, and processing the composite image 312 to produce a sham, focused image 314. By selecting an appropriate weighting sequence 308, the effects of a fluttering shutter can be synthesised, and negative and non-binary weights may be used. Video analytic functions 304, such as background subtraction, tracking and occlusion detection may be used to improve the results of the de-blurring 312. This image processing system 300 improves the quality of the input images for surveillance and biometric systems, including CFAIRS (combined face and iris recognition system) .