Target Extraction Using Gaussian Brightness Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing target extraction methods are limited in accurately identifying targets of varying sizes in images, as they rely on a single background image model and fail to account for changes in target size, leading to incomplete or incorrect target detection.
Innovation Solution
A method and apparatus that utilize Gaussian distribution modeling of pixel brightness values to generate input and background image modeling results, calculate likelihood ratios based on correlations with multiple target templates of different sizes, and determine target locations, enabling accurate target extraction regardless of size changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single background image model is used for target extraction, then the method is simple, but the target detection accuracy deteriorates when target size changes
Solution Approach 1:
The patent segments the target detection problem by dividing it into multiple size categories. It creates multiple background image models, each corresponding to a specific target size range. This segmentation allows the system to select the appropriate model based on the detected target size, thereby maintaining high detection accuracy across varying target sizes while keeping each individual model relatively simple.
Solution Approach 2:
The patent changes the parameter of background image model count from 1 to multiple models. By introducing multiple background image models with different size parameters, the system can adapt to targets of varying sizes. Each model is trained with brightness value distributions specific to its size category, enabling accurate target extraction regardless of target size changes.
2Measurement precision
If multiple target templates of different sizes are used, then target detection accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-calculating and storing brightness value distributions for multiple target sizes during the training phase. Multiple background image models are prepared in advance, each with pre-computed statistical characteristics. During actual target detection, the system only needs to compare input image brightness values against these pre-prepared models, significantly reducing real-time computational complexity while maintaining high detection accuracy.
3Adaptability or versatility
If Gaussian distribution modeling is applied to brightness values, then the ability to handle varying target sizes improves, but the computational requirements increase
Solution Approach 1:
The patent applies Gaussian distribution modeling to brightness values by changing the parameter representation from raw brightness values to Gaussian distribution parameters (mean and variance). This parameter transformation enables the system to efficiently model and compare brightness distributions across different target sizes. The Gaussian model provides a compact mathematical representation that reduces computational energy requirements compared to analyzing full brightness value distributions.
Data Source
AI summary
A target extracting apparatus and method for extracting a target through probability modeling of pixel brightness values in an image, and a recording medium storing a program for performing the method, are disclosed. The method includes extracting a background image from an input image, modeling brightness values of the input image and the background image based on a Gaussian distribution to generate an input image modeling result and a background image modeling result, calculating likelihood ratios for determining whether a pixel in the input image corresponds to a target based on a correlation between the input image modeling result and the background image modeling result and target templates having different sizes; determining a location of the target based on the calculated likelihood ratios; and outputting the determined location of the target.