Space-Time Video Slice Indexing for Fast Duplicate Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video document search systems face inefficiencies in processing large volumes of data, leading to slow and complex information processing, and struggle to reliably detect complete or partial copies of video documents in databases, especially in peer-to-peer networks and video sharing sites.
Innovation Solution
The method involves representing video documents as parallelepipedal volumes by juxtaposing frames on a time axis, constructing space-time video slices by extracting pixels at specific angles, normalizing dimensions, detecting contours and changes in planes, and creating a video fingerprint using characteristics vectors and binary decision trees for efficient indexing and comparison.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional video document search systems process large volumes of data, then they can search comprehensive databases, but processing speed becomes slow and complexity increases
Solution Approach 1:
The patent segments video documents into space-time slices by extracting pixels at specific angles (0°, 45°, or 90°) from successive frames. Each slice represents a temporal segment transformed into a 2D image format, enabling parallel processing and reducing the complexity of analyzing large video databases while maintaining comprehensive search capability.
Solution Approach 2:
The patent extracts only the necessary pixel information from each video frame at predetermined angles to create space-time slices. This extraction process removes redundant data while preserving essential visual characteristics, significantly reducing processing volume and enabling faster search operations across large video collections.
2Reliability
If conventional systems index video documents comprehensively, then search reliability improves, but calculation time increases
Solution Approach 1:
By dividing video documents into discrete space-time slices at specific angles, the system can independently process and index each slice, then combine results for comprehensive search. This segmentation maintains detection reliability for complete and partial copies while reducing overall calculation time through parallel processing of slices.
Solution Approach 2:
The patent performs preliminary extraction and indexing of space-time slices before actual search operations. By pre-processing video documents into standardized angular slices and storing them in an indexed format, the system enables rapid retrieval and comparison during search, significantly reducing calculation time while maintaining comprehensive detection capability.
3Adaptability or versatility
If the system processes video documents with various transformations, then adaptability improves, but processing complexity increases
Solution Approach 1:
The patent creates a universal indexing approach by extracting pixels at multiple predetermined angles (0°, 45°, 90°) from video frames. This multi-angle extraction generates space-time slices that can represent the same video content under different transformations, enabling the system to recognize and match videos regardless of rotational or angular variations while using a consistent processing framework.
Data Source
AI summary
A method of indexing a video document represents the video document in the form of a parallelepipedal volume obtained by juxtaposing, on a time axis, successive frames of the video document where each frame is represented by a matrix of pixels in a two-dimensional space. A space-time video slice of the video document is constructed by extracting a set of pixels for each frame of the video document in a same line defined relative to a first direction of the frames and juxtaposing the lines of pixels corresponding to the frames on a time axis to form a space-time video slice of a length defined by the juxtaposed lines of pixels and of a height equal to the length of the lines extracted from each of the frames. The length is a function of the dimension of the frames in a second direction and the orientation of the lines. The colorimetric values of the pixels of the lines of the space-time video slice are initialized on the basis of the content of the lines extracted from the frames of the video document.