Automated Video Analysis for Theft Detection in Retail

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current monitoring systems in retail stores are inefficient in promptly identifying stolen items, requiring extensive time to review recorded videos and often fail to capture the theft scene, leading to delayed or abandoned investigations.

Innovation Solution

An image processing apparatus that detects a person's behavior history from video footage and outputs information about stolen products, using a detection unit and output unit to quickly identify and report stolen items based on predetermined conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If video footage is manually reviewed to identify stolen items, then investigation thoroughness is improved, but investigation time increases significantly

Engineering Contradiction:

Improveinvestigation thoroughnessVSAvoidinvestigation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual video review (mechanical human operation) with an automated image processing system that uses detection units to extract person information and behavior histories, and output units to identify stolen items automatically. This substitution dramatically reduces investigation time while maintaining thoroughness through systematic automated analysis of video footage.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service by automatically processing video footage without requiring manual intervention. The detection unit autonomously extracts person information, the behavior history unit automatically analyzes behaviors, and the output unit independently identifies stolen items, allowing the system to serve itself in completing the investigation workflow.

Inventive Principle:

Principle #25Self-service

2Reliability

If comprehensive video review is performed to ensure no theft scene is missed, then detection reliability is improved, but operational complexity increases

Engineering Contradiction:

Improvedetection reliabilityVSAvoidoperational complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the complex video review process into distinct functional modules: a detection unit for extracting person information, a behavior history unit for analyzing behaviors, and an output unit for identifying stolen items. This segmentation maintains detection reliability through comprehensive analysis while reducing operational complexity by automating each segment with specialized processing functions.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The behavior history unit acts as an intermediary between raw video data and stolen item identification. It processes and structures behavior information, serving as a mediator that transforms complex video content into organized behavior histories that facilitate reliable and simplified theft detection.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Measurement precision

If manual verification of each stolen product is performed, then accuracy in identifying stolen items is improved, but productivity decreases

Engineering Contradiction:

Improveidentification accuracyVSAvoididentification speed

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual verification operations with automated image processing that uses detection units to extract person information and output units to identify stolen items. This substitution maintains identification accuracy through systematic analysis while dramatically improving productivity by eliminating time-consuming manual verification of each product.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

4Measurement precision

If detailed behavior analysis is performed on each person, then theft detection accuracy is improved, but processing time increases

Engineering Contradiction:

Improvetheft detection accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The behavior history unit performs self-service by automatically analyzing person behaviors without requiring manual intervention. It extracts behavior histories from video footage and processes them autonomously to identify theft-related actions, maintaining high detection accuracy while reducing processing time through automated systematic analysis.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system maintains continuous useful action by automatically processing behavior analysis without interruption. The detection unit continuously extracts person information, the behavior history unit continuously analyzes behaviors, and the output unit continuously identifies stolen items, ensuring both accuracy and efficiency through uninterrupted automated processing.

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS12260718B2Image processing apparatus, image processing method, and storage medium

Publication Date: 2025.03.25 CANON KK

AI summary

An image processing apparatus that outputs information about a specified product includes a detection unit configured to detect person information including a behavior history of a person detected from a video, and an output unit configured to output information about a product that meets a predetermined condition, based on the detected behavior history in the person information relating to a person to be retrieved.