License plate image generation method and related device

By using a license plate enhancement model based on artificial neural networks and a residual local feature network, the problem of clarifying blurry license plate image data was solved, and the clear display of license plate characters was achieved, thus improving the quality of license plate images captured by traffic cameras.

CN116311210BActive Publication Date: 2026-05-29HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HANGZHOU HIKVISION DIGITAL TECHNOLOGY CO LTD
Filing Date
2023-03-29
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Existing technologies are insufficient to effectively enhance and restore blurry license plate images into clear images. Traditional methods suffer from problems such as uneven or excessive lighting, blurry fonts, and decreased contrast.

Method used

A license plate enhancement model based on artificial neural networks is adopted. The training data is used to enhance both blurred and clear license plate image data. The model combines residual local feature networks and attention mechanisms to enhance and fuse the license plate region.

Benefits of technology

It improves the overall visual effect of license plate images, ensuring that the characters on the license plate are clearly visible, avoiding problems such as blurry fonts, overexposure, underexposure, and uneven background color, and does not require modification of traffic cameras, thus reducing acquisition costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116311210B_ABST
    Figure CN116311210B_ABST
Patent Text Reader

Abstract

Embodiments of the present application disclose a license plate image generation method and related equipment. The license plate image generation method comprises: extracting license plate image data containing a license plate region based on a collected license plate image, the image format of the license plate image being a bayer format; performing enhancement processing on the license plate image data using a license plate enhancement model to obtain a license plate region enhanced image, wherein the license plate enhancement model is a model obtained by training an artificial neural network based on a training data pair, the training data pair being composed of license plate blurred image data and license plate clear image data obtained based on historical collected license plate images; and fusing the license plate region enhanced image into the license plate region in the license plate image to generate a target license plate image after license plate enhancement. Through the present application, the technical problem that it is still difficult to enhance and restore a license plate clear image corresponding to license plate blurred image data is solved.
Need to check novelty before this filing date? Find Prior Art