An ultra-high-definition video noise reduction method

CN121010522BActive Publication Date: 2026-04-21GUANGDONG TUSHENG ULTRA HD INNOVATION CENT CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
GUANGDONG TUSHENG ULTRA HD INNOVATION CENT CO LTD
Filing Date
2025-08-20
Publication Date
2026-04-21

AI Technical Summary

Technical Problem

Traditional noise reduction algorithms struggle to balance noise reduction and detail preservation in low-light environments, especially in low-light videos where noise distribution is uneven and the signal-to-noise ratio is low, leading to blurred edges and loss of detail.

Method used

An orthogonal matching tracking algorithm based on scene region adaptation is adopted. By processing video frames in blocks, a dedicated dictionary is constructed based on the differences in grayscale and texture features, and differential encoding and noise reduction are performed. The DCT coefficients are optimized to preserve detailed information.

Benefits of technology

It achieves ultra-high-definition video noise reduction in low-light scenes, preserving video details to the maximum extent, and improving the practicality and accuracy of video analysis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121010522B_ABST
    Figure CN121010522B_ABST
Patent Text Reader

Abstract

This application relates to the field of video denoising technology, specifically to an ultra-high-definition video denoising method. The method includes: acquiring video data and noise variance; obtaining grayscale thresholds and gradient thresholds based on the grayscale values ​​and gradient values ​​of pixels in each frame of the video data, and acquiring the region of each frame image; then dividing the image into blocks by determining growth criteria based on grayscale differences and gradient differences; determining grayscale feature weights and texture feature weights based on the differences between grayscale values ​​and grayscale thresholds, and the differences between gradient values ​​and gradient thresholds; improving the original DCT coefficients based on the feature weights and texture feature weights to construct a dedicated dictionary; denoising, decoding, and reconstructing the image based on the dedicated dictionary; and synthesizing each reconstructed frame image into a complete video. This application achieves optimal denoising of video image frames while preserving the video's detailed information to the maximum extent.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Sonar image segmentation method and device based on NSCT domain denoising and self-adjusting spectral clustering

    CN117726634A

  • Security and protection monitoring video real-time transmission method based on Internet of Things

    CN120050421A