Video Occupant Segmentation for Real-Time Anomaly Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional surveillance systems in retail environments struggle to detect anomalous behavior among occupants in real-time due to computational complexity and inefficiency, posing security risks.

Innovation Solution

A system and method utilizing a video capture device and processing device to generate segmentation information, bounding boxes, and determine an Intersection over Union (IOU) region for bounding boxes to identify anomalous behavior, generating alerts for potential threats.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional video anomaly detection techniques are used to monitor retail environments, then security monitoring capability is provided, but computational complexity increases and real-time detection effectiveness deteriorates

Engineering Contradiction:

Improvesecurity monitoring capabilityVSAvoidcomputational complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the video feed into multiple frames and divides the monitoring task by processing only specific regions of interest (ROI) rather than the entire video stream. This segmentation approach reduces computational complexity while maintaining security monitoring capability, as the system analyzes only critical areas where anomalous behavior is likely to occur.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent extracts and processes only the most critical information from video frames, such as bounding boxes around potential occupants and their movement patterns, rather than processing the complete video data. This extraction method reduces computational complexity by focusing resources on the most relevant visual elements for security monitoring.

Inventive Principle:

Principle #2Taking out (Extraction)

2Area of stationary object

If manual monitoring of multiple CCTV cameras is performed to detect security risks, then comprehensive surveillance coverage is achieved, but operational difficulty and time consumption increase

Engineering Contradiction:

Improvesurveillance coverageVSAvoidmanual monitoring difficulty

Core Design Contradiction:

Area of stationary objectVSEase of operation

Solution Approach 1:

The patent implements an automated system that performs security monitoring independently without requiring manual intervention. The system automatically processes video feeds, detects anomalous behavior, and generates alerts, thereby eliminating the operational difficulty and time consumption associated with manual monitoring while maintaining comprehensive surveillance coverage.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent incorporates a feedback mechanism where the system continuously monitors video feeds, compares current frames against established patterns, and automatically adjusts its detection parameters. This feedback loop enables the system to adapt to changing environmental conditions while maintaining comprehensive surveillance coverage without manual intervention.

Inventive Principle:

Principle #23Feedback

3Speed

If real-time detection of anomalous behavior is implemented using conventional methods, then security responsiveness is improved, but processing time and computational resources increase

Engineering Contradiction:

Improvereal-time detection speedVSAvoidcomputational resources

Core Design Contradiction:

SpeedVSUse of energy by moving object

Solution Approach 1:

The patent applies partial action by processing only the necessary portions of video data required for anomaly detection, such as extracting bounding boxes and analyzing movement patterns in specific regions. This approach achieves real-time detection speed while reducing computational resource consumption by avoiding unnecessary processing of the entire video stream.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The patent performs preliminary processing of video frames to identify potential occupants and their locations before conducting detailed anomaly detection. This preliminary action, such as generating bounding boxes and calculating IOU regions, enables real-time detection by pre-organizing data in a way that facilitates rapid analysis of anomalous behavior with reduced computational resources.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12505638B2System and method to determine anomalous behavior

Publication Date: 2025.12.23 SENSORMATIC ELECTRONICS CORP

AI summary

The present disclosure relates to a system for determining anomalous behavior. The system comprises a processing device that is configured to generate first segmentation information and second segmentation information from the video feed, the first segmentation information corresponding to a first potential occupant and the second segmentation information corresponding to a second potential occupant. The processing device further configured to determine that the first segmentation information and the second segmentation information correspond to actual occupants based on the first segmentation information and the second segmentation information, and generate a first bounding box associated with the first segmentation information and a second bounding box associated with the second segmentation information. The processing device further configured to determine an intersection over union (IOU) region for the first bounding box and the second bounding box, identify anomalous behavior based on the IOU region, and generate one or more alerts identifying the anomalous behavior.