VORTEX: Video retrieval and tracking from compressed multimedia databases-visual search engine

Abstract

Multimedia data is generally stored in compressed form in order to eciently utilize the available storage facilities. Access to archives is dependent on our ability to browse compressed multimedia information| retrieval and tracking from coded video databases. In this paper, a novel visual search engine for video retrieval and tracking from compressed multimedia databases is proposed. The goal of the project is the implementation of a visual browser that operates in a distributed environment where users initiate video searches and retrieve relevant video information simultaneously from multiple video a r chives. Being presented with a query in the form of template images of objects, the system operates on the compressed video t o nd the images or video s e quences where those objects are present and their positions in the image. Upon user's request, the system will decompress and display only the video s e quences of interest.

VORTEX: Video Retrieval and Tracking from Compressed Multimedia Databases—Multiple Object Tracking from MPEG-2 Bit Stream

“…VORTEX emphasizes the concept of operating on compressed video data and exploiting the critical motion compensation information produced by the encoder's motion-based prediction for real-time object tracking and video retrieval [26][27][28][29]. The central idea is to intermittently initialize detection and tracking temporally by using a video parsing method and at the same time initialize spatially by using an available pattern recognition technique (template matching, affine invariant matching, use object indexing data or feature vector information).…”

Content-Based Image Retrieval by a Fuzzy Scale-Space Approach

“…4 Recently, systems that retrieve video sequences have emerged too, like the Columbia University's VisualSEEK, 53 or VORTEX, from the Illinois University at Chicago. 50 Here we focus on the use of fuzzy scale-space primitive features for the efficient realization of an image retrieval algorithm. In particular, we show how the algorithm reported in Ref.…”