About cookies on this site Our websites require some cookies to function properly (required). In addition, other cookies may be used with your consent to analyze site usage, improve the user experience and for advertising. For more information, please review your options. By visiting our website, you agree to our processing of information as described in IBM’sprivacy statement. To provide a smooth navigation, your cookie preferences will be shared across the IBM web domains listed here.
Abstract
In this paper, we present retrieval methods for extracting colored moving objects from surveillance cameras. In particular, we describe a method to classify moving objects into one of six colors. The method includes two sets of parameters. The first set can be used to compensate for illumination conditions and camera differences. The second set is used to tune the color extraction for specific object types and optimal retrieval. The tuning can be performed after video analysis has extracted metadata in real-time. The metadata is fed into a database which can be interactively queried by the user. The ability to query with feedback makes it possible to successfully find events of interest during an investigation without performing real-time adaptation which might be error-prone. In addition, for situations in which tracking moving objects is difficult, the system can generate significant colored object snippets of moving colored objects for retrieval. If the event of interest is not found by the initial system, this backup system can be invoked ensuring that the relevant event is found. © 2008 IEEE.