Time Scale Adaptive Motion Detection for Video Segments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motion detection techniques in video-based tracking applications are computationally expensive and fail to effectively handle objects moving at varying speeds or remaining stationary, leading to incorrect categorization of stationary objects as background.
Innovation Solution
A method and system for efficient, time-scale-adaptive video-based motion detection that evaluates video segments to identify pixel classes indicative of stationary and moving pixels, combining frame differencing and background estimation/subtraction to define a final motion mask for non-persistent objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional motion detection algorithms are tuned to support a limited range of speeds, then detection accuracy is improved for objects within that range, but detection reliability deteriorates when objects move at varying or inconsistent speeds
Solution Approach 1:
The patent implements dynamic time-scale adaptation by adjusting the time-scale parameter based on detected motion characteristics. The system transitions from static parameter tuning to dynamic parameter adjustment, allowing the motion detection algorithm to adapt to varying object speeds in real-time traffic scenarios, thereby maintaining both accuracy and reliability across different motion patterns
Solution Approach 2:
The patent changes the time-scale parameter of the motion detection algorithm based on detected motion patterns. By modifying the time-scale parameter dynamically rather than using fixed parameters, the system can effectively detect objects across a wide range of speeds, resolving the contradiction between precision for specific speeds and reliability across varying speeds
2Productivity
If motion detection algorithms are tuned for limited speed ranges, then computational efficiency is improved, but adaptability deteriorates when facing objects with varying motion patterns
Solution Approach 1:
The system employs dynamic time-scale adjustment that adapts to detected motion characteristics. This dynamic approach allows the algorithm to maintain computational efficiency by only adjusting parameters when necessary, while simultaneously improving adaptability to handle objects with varying speeds and motion patterns in traffic scenarios
Solution Approach 2:
The patent dynamically changes the time-scale parameter based on motion detection needs. This parameter adaptation enables the system to maintain computational efficiency through selective adjustment rather than continuous processing, while achieving versatile detection across diverse object speeds and motion patterns
3Measurement precision
If stationary objects are categorized as background, then background modeling accuracy is improved, but measurement precision deteriorates when stationary objects should be detected as foreground
Solution Approach 1:
The patent implements dynamic time-scale analysis to distinguish between stationary background and stationary foreground objects. By analyzing motion at multiple time-scales, the system can identify objects that remain stationary for extended periods versus true background elements, improving both background modeling accuracy and detection reliability of stationary objects of interest
Solution Approach 2:
The patent introduces a temporal dimension to the detection process by analyzing motion patterns across different time-scales. This additional temporal dimension allows the system to differentiate between stationary objects that should be detected (foreground) and true background elements, resolving the contradiction between accurate background modeling and reliable stationary object detection
Data Source
AI summary
A method and system for efficient non-persistent object motion detection comprises evaluating a video segment to identify at least two first pixel classes corresponding to a plurality of stationary pixels and a plurality of pixels in apparent motion, and evaluating the video segment to identify at least two second pixel classes corresponding to a background and a foreground indicative of the presence of a non-persistent object. The first pixel classes and the second pixel classes can be combined to define a final motion mask in the selected video segment indicative of the presence of a non-persistent object. An output can provide an indication that the object is in motion.