Cross-domain HDR image dynamic ghosting removing method fusing statistical offset fuzzy membership degree

By incorporating statistical offset fuzzy membership degrees, a cross-domain dynamic ghosting method for HDR images is proposed, which solves the problems of ghosting and motion blur in HDR imaging. It achieves pixel-level adaptive processing and optimized allocation of computing resources, thereby improving the quality and efficiency of image reconstruction.

CN120852252APending Publication Date: 2025-10-28NORTHWEST NORMAL UNIVERSITY
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202510951073.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-10
Publication Date
2025-10-28

AI Technical Summary

Technical Problem

Existing HDR imaging technologies face problems such as ghosting and motion blur caused by spatial inconsistencies when processing multi-exposure images. Furthermore, deep learning methods lack pixel-level fine control and adaptive allocation of computing resources, resulting in limited reconstruction quality and efficiency in complex scenes.

Method used

A cross-domain HDR image dynamic deghosting method that integrates statistical offset fuzzy membership is adopted. Through the spatial context alignment module, statistical offset fuzzy membership module and multi-scale cross-domain collaborative processing module, pixel-level adaptive processing and on-demand allocation of computing resources are achieved. The fuzzy membership function is used to quantify the degree of pixel deviation, and a membership map is generated for intelligent gating to activate computationally intensive modules.

Benefits of technology

It effectively reduces ghosting artifacts, preserves real scene details, and improves the visual quality and computational efficiency of HDR images, especially significantly improving reconstruction quality and robustness in complex scenes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120852252A_ABST
    Figure CN120852252A_ABST
Patent Text Reader

Abstract

The invention discloses a cross-domain HDR image dynamic ghosting removal method fusing statistical offset fuzzy membership, and the method comprises the steps: carrying out the ghosting removal of an LDR image sequence through employing a trained fuzzy statistical cross-domain ghosting removal model, and obtaining a ghosting-free HDR image; according to the model, after an input LDR sequence is subjected to preprocessing and feature extraction, a space context alignment module dynamically aligns non-reference frame and reference frame features through space-channel attention; meanwhile, the LDR image sequence is input into a statistical offset fuzzy membership module, and a membership graph is generated through sliding window statistical analysis and a fuzzy membership function; the aligned features are input into a multi-scale cross-domain cooperative processing module, multi-scale feature extraction and feature fusion are carried out, a frequency attention module is activated through membership graph gating, and global context modeling is introduced into a key region; and finally, outputting a high-quality ghosting-free HDR image. According to the method, fuzzy statistics and cross-domain processing are fused, and accurate ghosting removal and detail reservation in a dynamic scene are realized.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Cited By

  • Double-stage image restoration method for composite degradation

    CN121120451A

  • A two-stage image restoration method for complex degradation

    CN121120451B