License Plate Recognition via Edge Density Segmentation and Voting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional vehicle license plate recognition technologies face challenges in accurately recognizing license plates due to noise, varying illumination, and different viewing angles, leading to low recognition rates and slow computation speeds, especially when images contain complex backgrounds and similar features to the license plate.
Innovation Solution
A method utilizing edge-labeling technology to detect license plates by calculating edge densities and defining regions, followed by binarization and character recognition, which integrates recognition results through a voting technique to enhance accuracy and reliability, eliminating the need for extensive image training and improving noise tolerance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional binarization methods are used on the entire license plate region, then the process is simple, but noise and contamination significantly reduce recognition accuracy
Solution Approach 1:
The patent divides the license plate region into multiple character sub-regions individually. Each character region undergoes separate binarization and noise filtering processing, rather than processing the entire plate as one unit. This segmentation allows noise to be effectively removed from each character while maintaining overall recognition accuracy.
Solution Approach 2:
The patent extracts character-related data from the license plate image through histogram projection, separating the character information from the background. This extraction process isolates the useful character data from noisy background elements, enabling more accurate recognition.
2Productivity
If the license plate occupies a small area in the image, then the system can process images faster, but it becomes harder to obtain stable binarized data due to noise
Solution Approach 1:
By segmenting the small license plate region into individual character components, the patent maintains processing efficiency while improving binarization stability. Each character can be processed independently with optimized parameters, reducing the impact of noise even when the overall plate area is small.
Solution Approach 2:
The patent applies different processing parameters and noise filtering strategies to different character regions based on their local characteristics. This local quality approach ensures that each character receives appropriate processing attention, maintaining stability even when the license plate occupies a small image area.
3Measurement precision
If extensive image training is performed to improve recognition accuracy, then recognition rate increases, but computation time and system complexity increase significantly
Solution Approach 1:
The patent employs noise filtering and character segmentation algorithms that automatically adapt to the input image characteristics without requiring external training data. The system self-adjusts processing parameters based on local image features, eliminating the need for time-consuming training phases while maintaining high recognition rates.
Solution Approach 2:
The patent dynamically adjusts processing parameters such as binarization thresholds and noise filtering strength based on the actual image conditions. This parameter adaptation allows the system to achieve high accuracy without fixed training, as parameters are optimized in real-time for each image.
4Reliability
If conventional recognition methods are used that process the entire license plate as one unit, then the system is simpler to implement, but it is highly susceptible to noise and contamination affecting overall recognition
Solution Approach 1:
The patent divides the license plate into individual character regions and processes each separately. This segmentation ensures that noise or contamination in one character region does not propagate to affect the recognition of other characters, significantly improving overall noise tolerance.
Solution Approach 2:
The patent applies noise filtering that identifies and discards contaminated or noisy character regions, then recovers or corrects the recognition results based on the filtered data. This approach maintains reliability by eliminating the impact of noise while preserving the overall recognition process.
Data Source
AI summary
A vehicle license plate recognition method and a system thereof are disclosed. A region where a vehicle license plate image exists is detected according to the edge densities of an input image and a vehicle license plate specification. A text area of the vehicle license plate image is divided into a plurality of character images. The character images are binarized to obtain a plurality of binarized character images. A plurality of characters is recognized from the binarized character images. The characters are recombined to form a character string. The abovementioned steps are repeated to obtain a new character string from another image of the same vehicle, which is captured at a next time point. The character string is compared with the new character string character by character to obtain a comparison result for verifying reliability of recognition through a voting technique.