Multiple Gaussian Models for Image Background Maintenance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing moving-object detection techniques face misinterpretation issues due to foreground overstaying or marginal color similarity with the background, leading to incorrect background updates in changing environments.
Innovation Solution
A method using multiple Gaussian models, where a primary Gaussian model is established and updated based on the comparison of secondary Gaussian models, ensuring accurate background maintenance by distinguishing between background and foreground information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single Gaussian model is used to maintain the background, then the device complexity is low, but the measurement precision of background maintenance deteriorates due to misinterpretation when foreground overstays or marginal color approximates background color
Solution Approach 1:
The patent divides the background maintenance task into multiple Gaussian models (first Gaussian model for initial background, second Gaussian model for updated background) to segment the learning process. This segmentation allows the system to compare different background representations and avoid misinterpretation of foreground as background, thereby improving measurement precision without excessive complexity increase.
Solution Approach 2:
The patent introduces a temporal dimension by comparing background models across different time points (first background vs. second background). By analyzing changes in the background over time and using multiple Gaussian distributions to represent different background states, the system achieves more precise background maintenance while distinguishing it from foreground objects.
2Adaptability or versatility
If background learning is continuously updated to adapt to changing background, then the adaptability improves, but the reliability deteriorates due to foreground being learned and incorporated into background causing misdetermination
Solution Approach 1:
The patent performs preliminary comparison between the first Gaussian model (initial background) and the second Gaussian model (updated background) before finalizing background updates. By预先 comparing these models and determining whether changes represent actual background evolution or foreground intrusion, the system maintains adaptability to genuine background changes while preventing foreground misincorporation, thus preserving reliability.
Solution Approach 2:
The patent implements a feedback mechanism where the comparison result between first and second Gaussian models feeds back into the background update decision. If the comparison indicates that changes are due to foreground objects (misdetermination), the system corrects the background model accordingly. This feedback loop ensures both adaptability to real background changes and reliability in foreground detection.
Data Source
AI summary
A method maintaining an image background by multiple Gaussian models utilized to a device includes the following steps. First, the device captures an image frame having pixels to obtain background information, and then calculates the background information to establish a primary Gaussian model. Next, the device captures continuous image frames in a time period to obtain and calculate graphic information for establishing a secondary Gaussian model, and then repeates the steps to establish multiple secondary Gaussian models. Finally, the device compares two secondary Gaussian models, and then updates learning for the primary Gaussian model by the secondary Gaussian model if the graphic information of the secondary Gaussian models are attributable to the background information, or maintains the background information of the primary Gaussian model without updating the learning if anyone of the graphic information of the two secondary Gaussian models is unattributable to the background information.