License Plate Localization via Texture and Connected Component Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ALPR systems face challenges in accurately locating license plates due to poor image quality, image perspective distortion, and interference characters, which are not effectively addressed by prior art approaches relying on shape, texture, and color features.
Innovation Solution
A method and system that combines top-down texture analysis and bottom-up connected component analysis to locate license plates, involving image binarization, run classification, spatial clustering, and linearity testing to filter out non-text objects and identify plate edges, with high-pass filtering to adapt to different background colors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If prior art approaches use shape, texture, and color features to locate license plates, then the system can identify potential plate regions, but the accuracy is poor due to ambient lighting conditions, image perspective distortion, and interference characters
Solution Approach 1:
The patent segments the license plate detection process into two distinct phases: top-down texture analysis to identify potential plate regions, and bottom-up connected component analysis to locate text characters. This segmentation allows each phase to specialize in specific tasks, improving overall accuracy while handling adverse conditions through targeted feature analysis in each segment
Solution Approach 2:
The patent changes the analytical parameters used for detection by examining texture patterns, run-length characteristics, and spatial relationships rather than relying solely on traditional shape, texture, and color features. This parameter transformation enables the system to distinguish license plates from interference characters even under poor lighting and perspective distortion
2Measurement precision
If the system performs comprehensive image analysis to accurately locate license plates, then the localization accuracy improves, but the computational effort and processing time increase
Solution Approach 1:
The patent performs preliminary action by conducting top-down texture analysis first to identify potential license plate regions before performing the more computationally intensive bottom-up connected component analysis. This preliminary filtering reduces the search space and minimizes computational effort in subsequent processing steps
Solution Approach 2:
By dividing the analysis into two sequential phases (top-down then bottom-up), the patent avoids performing comprehensive analysis on the entire image simultaneously. Each phase processes only relevant portions, reducing overall computational complexity while maintaining high localization accuracy
3Ease of manufacture
If the system uses traditional feature-based detection methods, then the implementation is relatively simple, but the system produces false alarms and cannot handle varying background colors effectively
Solution Approach 1:
The patent changes the detection parameters from traditional shape and color features to texture-based features including run-length analysis and spatial clustering metrics. This parameter transformation enables effective handling of varying background colors and reduces false alarms while maintaining implementation feasibility through established image processing techniques
Data Source
AI summary
A license plate localization method and system based on a combination of a top-down texture analysis and a bottom-up connected component. An image with respect to a vehicle captured by an image capturing unit can be processed in order to locate and binarize a busy area. A black run with respect to the binarized image can be analyzed and classified and one or more objects (connected components) can be generated based on the black run classification. The objects can be further classified in accordance with their size utilizing a run-length based filter to filter out a non-text object. The leftover objects can then be spatially clustered and the uniformity and linearity of the clustered objects can be examined based on a linearity test. The clustered objects can be rejected if they fail the linearity test and the detected objects can further be matched with a plate edge characteristic in order to locate a license plate.