Camera Device Luminance Signal Analysis for Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional camera monitoring systems fail to automatically detect video-image abnormalities caused by camera device failures, such as dark images, which can be misinterpreted as environmental conditions rather than device malfunctions, especially in large-scale systems where visual observation is impractical and delays incident detection.
Innovation Solution
A monitoring system utilizing IP cameras connected to a server and recording device, with a measurement unit that employs algorithms to detect abnormalities in luminance signals, including changes in mean values and flickering patterns, to differentiate between environmental changes and camera failures, allowing for automatic detection and notification of issues without human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual observation by an observer is used to detect video-image abnormalities, then the observer can comprehensively determine abnormalities by empirically understanding the photographing environment, but it is difficult to visually and immediately detect abnormalities of all camera devices in large-scale systems and is not suitable for small-scale systems requiring manpower savings
Solution Approach 1:
The patent replaces the mechanical/visual observation system with an automated electronic detection system. The abnormality detection unit automatically analyzes video signals from multiple cameras using electronic processing algorithms, substituting human visual inspection with machine-based signal analysis that can handle large-scale systems efficiently while maintaining detection accuracy
Solution Approach 2:
The system enables self-diagnosis of camera devices through automated abnormality detection. Each camera's video signal is automatically analyzed by the detection unit, allowing the system to self-monitor and identify abnormalities without requiring external human observation, thus improving both efficiency and scalability
2Productivity
If conventional automatic detection methods are used, then some failures can be detected, but cannot automatically detect all abnormalities in image data such as dark images that may be misinterpreted as environmental conditions
Solution Approach 1:
The abnormality detection unit continuously monitors video signals and provides feedback about detected abnormalities. The system compares current video signal characteristics against expected ranges and provides automated feedback when deviations are detected, enabling continuous improvement of detection accuracy through systematic analysis rather than single-point detection
Solution Approach 2:
The patent employs multiple detection parameters including luminance signal levels, signal presence/absence, and temporal changes in video characteristics. By monitoring multiple parameters simultaneously, the system can distinguish between environmental variations (such as legitimate dark scenes) and actual camera failures, improving detection precision while maintaining automated operation
3Productivity
If recording video-images for later checking is used in unmanned store surveillance, then manpower is saved during normal operation, but video-image abnormalities are not detected until the recorded video-image is visually checked at the time of incident occurrence
Solution Approach 1:
The abnormality detection unit operates continuously in the background, constantly analyzing video signals from all cameras without interruption. This continuous automated monitoring ensures that abnormalities are detected immediately as they occur, eliminating detection delays while maintaining the unmanned operation model and manpower efficiency
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In a monitoring system, each individual camera device divides image data acquired by an image sensor unit into a plurality of areas, and detects video-image abnormality from an average luminance signal value in each area or from a change thereof. The camera device that has detected video abnormality transmits video-image abnormality detection information to a server. In the plurality of areas, it is possible to use different detection algorithms such as a first algorithm in which it is determined that an abnormality has occurred when there is no change in the average luminance value, a second algorithm in which it is determined that an abnormality has occurred when there is a change in the average luminance value, and a third algorithm for observing a change in the video captured while a change is made to an aperture, exposure time, or the like.