Motion Detection via Low Rank Matrix Frequency Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motion detection methods in surveillance systems require high computational complexity, making them inefficient for real-time processing, especially when dealing with large images, as they need to analyze numerous frames to detect motion accurately.
Innovation Solution
A method and system that utilize a low rank matrix capable of performing frequency domain transform, filtering, and time domain transform in a single operation to detect motion by generating a time domain matrix from pixel value variations and filtering out low-frequency components, reducing calculation complexity through centro-symmetry and fixed points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional motion detection methods (temporal difference, background subtraction, Gaussian modeling) are used to detect motion in surveillance images, then motion detection accuracy is maintained, but calculation complexity becomes too high for real-time processing
Solution Approach 1:
The patent transforms the motion detection problem from spatial domain to frequency domain by applying 2D-DCT (two-dimensional discrete cosine transform). This parameter change in the mathematical domain allows motion detection to be performed by comparing frequency coefficients rather than processing entire image frames, dramatically reducing computational complexity while preserving motion detection accuracy
Solution Approach 2:
The patent extracts only the essential motion information by comparing frequency coefficients (specifically DC and AC coefficients) between consecutive frames. Instead of analyzing all pixel data in each frame, the method extracts and compares only the relevant frequency components that contain motion information, reducing the amount of data to be processed
2Manufacturing precision
If conventional motion detection methods process images at resolution of 640×840, then sufficient detail is captured, but the processing speed drops to about 6 frames/sec which is insufficient for real-time surveillance
Solution Approach 1:
By changing the domain from spatial to frequency domain using 2D-DCT, the patent enables efficient processing of high-resolution images. The frequency domain representation compactly encodes image information, allowing fast comparison and motion detection even at 640×840 resolution, thereby achieving real-time processing speeds
Solution Approach 2:
The patent segments the image processing task by applying 2D-DCT to divide the image into frequency components (DC coefficient representing average intensity and AC coefficients representing variations). This segmentation allows independent processing and comparison of different frequency components, improving processing efficiency
Data Source
AI summary
A method and system for detecting motion are provided. The method includes: generating a time domain matrix including vectors corresponding to variation of pixel values as elements of the time domain matrix, of a video image including a plurality of frames; generating a motion matrix from which a low frequency area of the video image is removed by multiplying the time domain matrix by a low rank matrix; and generating a result image including a plurality of frames in which vectors, which are elements of the motion matrix, are included as variation of motion pixel values.