Surveillance Image Analysis Metadata Indexing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional surveillance systems require extensive image analysis for each image data set to access desired object data, leading to inefficiencies in monitoring and object detection.

Innovation Solution

A system comprising monitoring cameras, devices with image storage and analysis units, and a network that performs image analysis and metadata generation, allowing for efficient storage and retrieval of image data by using appropriate image analysis engines based on shooting conditions and target objects, facilitating quick access to specific image data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If image data is simply stored without analysis, then storage capacity is maintained, but image analysis must be performed on each image data set every time to access desired object data, leading to inefficiency

Engineering Contradiction:
Improveimage data access efficiencyVSAvoidtime for image analysis
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system performs image analysis in advance when image data is stored, extracting object information and generating metadata before the data is needed. This preliminary action creates an index structure that enables rapid retrieval without re-analyzing the full image data, directly resolving the contradiction between maintaining storage simplicity and achieving efficient access.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The invention introduces metadata as an intermediary between the stored image data and the retrieval system. This metadata contains extracted object information that acts as a bridge, allowing the system to access desired objects without performing complete image analysis on each query, thus improving productivity while reducing analysis time.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If extensive image analysis is performed on each image data set, then object detection accuracy is improved, but surveillance efficiency deteriorates due to the time-consuming nature of analyzing every image

Engineering Contradiction:
Improveobject detection accuracyVSAvoidsurveillance efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system segments the image analysis process into two stages: a comprehensive analysis stage that performs detailed object detection when data is stored, and a rapid retrieval stage that uses pre-generated metadata for quick access. This segmentation allows the system to maintain high detection accuracy through thorough initial analysis while achieving high surveillance efficiency through fast metadata-based retrieval.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

By performing complete image analysis and extracting object information in advance during the data storage phase, the system creates a ready-to-use metadata index. This preliminary action ensures that when surveillance operations occur, the system can rapidly query and retrieve relevant data without repeating the extensive analysis process, thereby maintaining both accuracy and efficiency.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If multiple image analysis engines are used to handle different shooting conditions and target objects, then analysis accuracy is improved, but system complexity increases

Engineering Contradiction:
Improveimage analysis accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system employs a single image analysis engine that is designed to handle multiple shooting conditions and target object types through configurable parameters and algorithms. This universal engine can adapt to different analysis requirements without requiring separate dedicated engines for each condition, thereby maintaining high analysis accuracy across diverse scenarios while avoiding the complexity of managing multiple specialized systems.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3843383B1Apparatus, system, method and program
Publication Date: 2023.10.18 YOKOGAWA ELECTRIC CORP
  • EP3843383B1 patent drawingFigure 1
  • EP3843383B1 patent drawingFigure 2
  • EP3843383B1 patent drawingFigure 3

AI summary

[Problem] If the shot image data is simply stored, image analysis must be performed on each of the image data each time in order to access the image data which includes a desired object. [Solution] A device is provided, comprising: an image storage unit which is configured to store image data obtained by a monitoring camera; an analyzing unit which is configured to perform image analysis on the image data and, in response to a detection of a target object to be detected among objects which are possibly included in the image data, generate feature data indicating a feature of the target object; and a transmitting unit which is configured to transmit metadata including the feature data and at least one of a shooting time of the image data in which the feature data is generated or an address of the image data within the image storage unit to a file server which is accessible to another apparatus different from the device.