Histogram Normalization for Accurate Frame Interruption Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance camera systems inaccurately determine interruptions due to changes in brightness, which can lead to false alarms and missed sharp changes in frames.
Innovation Solution
An image processing device that generates and normalizes histograms of pixel values from previous and current frames, calculates the similarity of these histograms, and sets a threshold value based on moment differences such as skewness and kurtosis to accurately determine interruptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If normalization is performed to match average values of histograms, then brightness changes can be prevented from being erroneously determined as interruptions, but the shape of the histogram is actually slightly changed according to a change in brightness, leading to inaccurate determination of sharp changes in frames
Solution Approach 1:
The patent changes the normalization parameter from matching average values to matching variation degrees (standard deviations) of pixel values. This parameter change resolves the contradiction by preserving the histogram shape characteristics while still compensating for brightness changes, thereby maintaining both interruption detection reliability and histogram shape similarity measurement precision.
2Ease of operation
If a fixed threshold value is used for similarity determination, then the processing is simple, but it cannot accurately adapt to different moment differences caused by brightness changes
Solution Approach 1:
The patent makes the threshold value dynamic by setting it based on the moment difference between histograms. Instead of using a fixed threshold, the system calculates the moment difference and uses it to determine an appropriate threshold, allowing the threshold to adapt to different brightness conditions while maintaining accurate interruption detection.
Solution Approach 2:
The patent changes the threshold parameter from a fixed value to a value that varies according to the moment difference. This allows the threshold to automatically adjust to different brightness conditions, resolving the contradiction between operational simplicity and determination accuracy.
3Productivity
If only average value normalization is performed, then the processing speed is maintained, but false alarms occur due to brightness changes being misinterpreted as interruptions
Solution Approach 1:
The patent changes the normalization parameter from average value matching to variation degree matching. This single parameter change maintains processing speed while eliminating false alarms caused by brightness changes, as it preserves the histogram shape information that distinguishes actual interruptions from brightness variations.
Data Source
AI summary
To accurately determine whether there is a sharp change in a frame.Provided is an image processing device including: a histogram generating unit configured to generate a previous histogram showing a distribution of pixel values in a previous frame that is generated before a predetermined frame and a current histogram showing a distribution of pixel values in the predetermined frame; a normalizing unit configured to perform normalization to match variations of the pixel values of the previous histogram and the current histogram; and a similarity determining unit configured to acquire a degree of similarity of shapes of the previous histogram and the current histogram after the normalization and determine whether the degree of similarity is greater than a predetermined similarity determining threshold value.