Fast non-local mean denoising method, device and equipment and readable storage medium
By combining the point to be denoised with the reference point in the nonlocal mean denoising method, the true and approximate similarity of the neighborhood block is calculated, which solves the problem of excessive time consumption in the existing technology and achieves efficient image denoising processing.
CN118096575BActive Publication Date: 2026-07-03WUHAN GUIDE INFRARED CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- WUHAN GUIDE INFRARED CO LTD
- Filing Date
- 2024-03-20
- Publication Date
- 2026-07-03
AI Technical Summary
Technical Problem
Existing nonlocal means denoising methods cannot significantly reduce computation time while ensuring denoising effect.
Method used
By combining the point to be denoised with the reference point to form a transitive combination, the true similarity and approximate similarity of the neighborhood block are calculated, reducing the number of times the true similarity of non-associated points in the search area is calculated, and a weighted average is used to calculate the output value.
Benefits of technology
It significantly reduces computation time without compromising noise reduction, thereby improving image processing efficiency.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN118096575B_ABST
Abstract
The application provides a fast non-local mean denoising method, device and equipment and readable storage medium, which comprises the following steps: for each to-be-denoised point in an original image, combining each to-be-denoised point with each reference point to form a transfer combination; for each transfer combination, calculating the real similarity of the neighborhood block of two pixel points; determining a current point and a search region; for each non-associated point in the search region, if the non-associated point and any to-be-denoised point form a transfer combination, determining a transfer path, calculating an approximate similarity according to the real similarity of each transfer combination involved in the transfer path, or calculating the real similarity; performing weighted average calculation on the pixel values of the pixel points in the search region to obtain the output value of the current point; and outputting a denoised image based on the output value of each to-be-denoised point. Through the application, the number of real similarity calculations is greatly reduced, thereby ensuring the denoising effect while significantly reducing the time consumption.
Need to check novelty before this filing date? Find Prior Art