Scene Change Detection Using Histogram and Flow Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera systems that detect scene changes and motion often transmit unnecessary data, leading to increased network bandwidth consumption and remote processing resource usage, as they fail to differentiate between significant and insignificant changes.
Innovation Solution
A system that employs Differential Scene Detection, Gaussian Scene Detection, and Flow Motion Detection processes to analyze YUV histograms and motion vectors, selectively streaming video data only when significant changes are detected, using a combination of hardware and software components like premotion estimators and video preprocessors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If camera systems transmit all detected motion data, then motion detection coverage is improved, but network bandwidth consumption increases
Solution Approach 1:
The patent segments motion detection into multiple analysis stages: initial frame comparison, histogram analysis of YUV color space distributions, and selective full transmission. Only frames showing significant scene changes (exceeding threshold deviations in histogram comparisons) are transmitted, while minor motions are filtered locally. This segmentation allows comprehensive motion detection while minimizing unnecessary data transmission.
Solution Approach 2:
The system performs preliminary motion analysis locally at the camera端 before transmission using histogram comparison and scene change detection algorithms. By pre-processing and filtering motion data locally, the system identifies and transmits only significant events, reducing overall network bandwidth consumption while maintaining reliable motion detection coverage.
2Measurement precision
If camera systems process and transmit all video data, then scene change detection accuracy is improved, but remote processing resource usage increases
Solution Approach 1:
The patent extracts and transmits only the essential scene change information (histogram deviation data, motion vectors indicating significant changes) rather than complete video frames. By taking out only the critical detection data needed for accurate scene change identification, the system reduces remote processing requirements while maintaining detection accuracy.
Solution Approach 2:
The system applies partial processing by performing complete scene change analysis locally at the camera and transmitting only when changes exceed predefined thresholds. This partial action approach avoids excessive transmission of unnecessary data while ensuring accurate detection of significant scene changes through thorough local analysis.
3Speed
If camera systems use simple motion detection, then processing speed is improved, but ability to differentiate significant changes deteriorates
Solution Approach 1:
The patent applies different processing qualities to different types of motion events. Simple frame comparison and histogram analysis are used for rapid initial screening of all frames, while more sophisticated scene change detection algorithms are applied selectively only to frames showing significant deviations. This local quality differentiation maintains high processing speed for routine monitoring while enabling accurate significance differentiation for notable events.
Data Source
AI summary
Techniques are generally described for scene change detection. A first and second histogram representing frames of a video may be received. A Euclidean distance between the first and second histogram may be determined. A third histogram of a third frame may be received. Values of the third histogram may be compared to corresponding values of a background model of the environment. A fourth frame and a fifth frame of image data of the video may be received. A flow value between a first block of pixels of the fourth frame and a corresponding second block of pixels of the fifth frame may be determined. The flow value may be determined based on a motion vector and a sum of absolute differences between the first and second block. A portion of the video may be streamed to a remote computing device if the flow value exceeds a threshold value.