License Plate OCR Correction via Hash Database

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies for recognizing vehicle license plates using optical character recognition (OCR) often suffer from errors, especially under unfavorable lighting conditions, which can lead to incorrect identification and potential misuse, such as toll evasion.

Innovation Solution

A system and method that utilize a hash function to generate a hash value from the OCR-recognized license plate, which is then compared to a correction database containing hash values and corresponding correction instructions. This allows for automatic correction of OCR errors without storing the actual license plate, thus maintaining data protection compliance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If OCR is used to recognize license plates, then automated identification is achieved, but recognition errors occur under unfavorable lighting conditions

Engineering Contradiction:
Improveautomated license plate identificationVSAvoidrecognition accuracy
Core Design Contradiction:
Extent of automationVSReliability

Solution Approach 1:

The system performs preliminary actions by capturing multiple images of the license plate at different locations and times before final recognition. These pre-captured images are stored and can be used to correct OCR errors if the initial recognition fails, thereby improving reliability while maintaining automation.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms where OCR recognition results are verified against stored images and previous recognition data. If errors are detected, the system uses feedback information to correct the recognition result, improving accuracy without requiring manual intervention and thus maintaining automated operation.

Inventive Principle:
Principle #23Feedback

2Reliability

If correction databases store actual license plates for error correction, then recognition accuracy improves, but data protection compliance is compromised

Engineering Contradiction:
Improveerror correction capabilityVSAvoiddata protection risks
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

Instead of storing actual license plate data in the correction database, the system creates and stores only hash values (digital fingerprints) of the license plates. These hash values are used for comparison and error correction while the actual license plate data remains protected and is not stored, thus maintaining both correction capability and data protection compliance.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The system introduces hash values as an intermediary between the license plate data and the correction database. The hash values serve as mediators that enable error correction through comparison while preventing direct storage of sensitive license plate information, thus resolving the contradiction between correction accuracy and data protection.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If multiple OCR readings are performed to improve accuracy, then error rate decreases, but processing time increases

Engineering Contradiction:
Improverecognition accuracyVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary OCR readings and stores the results along with confidence measures in a database. When a new license plate image is processed, the system quickly compares the new reading against pre-stored results using hash values, allowing for rapid error correction without requiring multiple sequential OCR readings, thus improving accuracy while minimizing processing time.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3869396B1Method and system for adjusting a label
Publication Date: 2025.04.23 TOLL COLLECT GMBH
  • EP3869396B1 patent drawingFigure 1~2
  • EP3869396B1 patent drawingFigure 3
  • EP3869396B1 patent drawingFigure 4

AI summary

A computer-implemented method and a system for correcting the license plate of a motor vehicle are disclosed.The procedure comprises the following steps: a) capturing (10) a license plate, b) generating (20) an OCR license plate from the captured license plate using text recognition, c) performing a reference test for the generated OCR license plate, the following steps being carried out only if the reference test is negative, d) forming (40) a hash value from the generated OCR license plate, e) comparing (50) the formed hash value with a correction database, the correction database comprising stored hash values ​​of license plates and correction instructions, each stored hash value being associated with at least one correction instruction, and f) if the formed hash value corresponds to a hash value stored in the correction database, applying (60) the at least one correction instruction associated with the stored hash value to the OCR license plate and generating a corrected OCR license plate.