Periodic Motion Detection for Multi-Grabbing Surveillance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current intelligent video surveillance systems face difficulties in tracking individual targets in crowded areas due to occlusions, limiting their ability to detect suspicious activities like shoplifting or cheating, which results in significant retail losses and security challenges.
Innovation Solution
A video-based surveillance system employing algorithms for detecting periodic motion patterns, specifically designed to identify 'multi-grabbing' behavior by analyzing motion data from surveillance videos, using components like change detectors, motion detectors, and periodic motion analyzers to generate alerts for real-time monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional IVS systems track individual targets in crowded areas, then target tracking capability is improved, but target occlusions cause tracking failure and detection accuracy deteriorates
Solution Approach 1:
The patent segments the target into multiple parts (head, body, limbs) and tracks each part separately using body part detection. This allows the system to maintain tracking reliability even when parts of the target are occluded by other objects or people in crowded areas, resolving the contradiction between tracking capability and detection accuracy.
Solution Approach 2:
The system performs periodic detection and analysis of motion patterns at regular intervals to identify suspicious behaviors such as multi-grabbing. By continuously monitoring and analyzing periodic motions, the system maintains reliable tracking and accurate detection even in dynamic crowded environments where occlusions occur intermittently.
2Reliability
If IVS systems monitor suspicious activities like shoplifting in real-time, then security detection capability is improved, but system complexity and computational requirements increase
Solution Approach 1:
The system focuses on detecting periodic motion patterns rather than analyzing all possible behaviors continuously. By identifying repetitive suspicious actions such as multiple grab-release cycles, the system achieves reliable security detection with reduced computational complexity compared to comprehensive real-time analysis of all activities.
Solution Approach 2:
The system changes the detection parameters from comprehensive target analysis to specific motion pattern recognition. By monitoring key parameters such as periodicity, motion amplitude, and temporal patterns, the system simplifies the detection process while maintaining high security detection capability for suspicious activities.
3Adaptability or versatility
If IVS systems distinguish multiple target behaviors in crowded scenes, then behavior recognition capability is improved, but target occlusions limit the types of behaviors that can be distinguished
Solution Approach 1:
By segmenting and tracking individual body parts separately, the system can distinguish different behaviors (such as normal browsing vs. multi-grabbing shoplifting) even when targets are partially occluded. This segmentation approach maintains behavior recognition versatility and distinction accuracy in crowded scenes where full target visibility is not always possible.
Data Source
AI summary
A method of video surveillance may include performing on input video at least one of the operations selected from the group consisting of motion detection and change detection, recording a motion pattern based on a result of said at least one of the operations, and analyzing the motion pattern to detect periodic motion in the video. A video surveillance apparatus may include a change detector, a motion detector, and/or a combination motion/change detector, a pattern analyzer, and a periodic motion detector.