High-precision detection of homogeneous object activity in a sequence of images

摘要

A first image is aligned with a second image. A first motion value is computed based at least in part on a sum of differences between corresponding pixels in the first image and the second image. A second motion value is computed further based at least in part on a third image. An impact score is generated based at least in part on a difference between values derived from the first motion value and the second motion value and an action is performed depending at least in part on whether the impact score indicates that damage has occurred among objects represented in the first image and the second image.