Exploring video streams using slit-tear visualizations

Abstract

Slit-tear visualizations allow users to selectively visualize pixel paths in a video scene. The slit-tear visualization technique is a generalization of the traditional photographic slit-scanning and more recent video slicing techniques: after a user specifies a pixel path of interest, the system generates a timeline that replicates those pixels for each frame in the video. These rich visualizations of the video data help users to discover and explore spatio-temporal patterns of activity in a video. In this video, we illustrate the use of slit-tear visualizations to detect movement and incidence of activity in a video scene, accentuate directional motion and small changes in the video, and discover patterns of activity between spatially distinct areas of the scene.