License Plate Character Segmentation via Likelihood Maximization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Automatic License Plate Recognition (ALPR) systems face challenges in recognizing diverse license plate layouts, distinguishing characters from obscuring objects, and handling varying imaging conditions, leading to segmentation errors and reduced accuracy.
Innovation Solution
A method and apparatus that generate models of license plate layouts by segmenting training images to extract characters and logos, calculating feature parameters, and estimating likelihood functions to compare with observed plates, reducing segmentation errors and improving accuracy by determining the closest matching layout configuration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional character segmentation methods are used, then the system can process license plates quickly, but the accuracy is reduced due to over- and under-segmentation errors
Solution Approach 1:
The patent applies segmentation by dividing the license plate image into multiple candidate regions and then selecting the most likely segmentation using likelihood functions. This allows the system to evaluate multiple possible character boundaries and choose the correct one, improving accuracy without significantly sacrificing processing speed.
Solution Approach 2:
The patent changes parameters by using likelihood functions that evaluate multiple segmentation hypotheses with different parameters (character widths, heights, positions). By comparing these parameter variations against learned models, the system achieves higher segmentation accuracy while maintaining efficient processing through mathematical optimization.
2Adaptability or versatility
If the system uses multiple layout models to handle diverse plate configurations, then adaptability improves, but device complexity increases
Solution Approach 1:
The patent implements universality by creating a unified likelihood-based framework that can handle multiple license plate layouts and configurations. Instead of implementing separate specialized systems for each layout type, a single multi-functional model evaluates all possible layouts using the same probabilistic approach, improving adaptability while managing complexity.
Solution Approach 2:
The patent uses an intermediary likelihood function as a mediator between the image data and multiple layout models. This intermediary layer probabilistically evaluates which layout configuration best explains the observed character patterns, allowing the system to handle diverse layouts without directly managing the complexity of each individual layout model.
3Measurement precision
If the system applies strict segmentation criteria to avoid errors, then accuracy improves, but the number of false rejections increases
Solution Approach 1:
The patent implements feedback through likelihood functions that evaluate segmentation hypotheses and provide probabilistic feedback on their validity. The system uses this feedback to iteratively refine segmentation decisions, adjusting criteria based on how well each hypothesis explains the observed data. This feedback mechanism allows the system to maintain high accuracy while reducing false rejections by accepting hypotheses with sufficient probabilistic support.
Data Source
AI summary
A method determines a license plate layout configuration. The method includes generating at least one model representing a license plate layout configuration. The generating includes segmenting training images each defining a license plate to extract characters and logos from the training images. The segmenting includes calculating values corresponding to parameters of the license plate and features of the characters and logos. The segmenting includes estimating a likelihood function specified by the features using the values. The likelihood function measures deviations between an observed plate and the model. The method includes storing a layout structure and the distributions for each of the at least one model. The method includes receiving as input an observed image including a plate region. The method includes segmenting the plate region and determining a license plate layout configuration of the observed plate by comparing the segmented plate region to the at least one model.