计算机视觉实时图像增强系统
By constructing a closed-loop system that integrates scene perception, enhanced scheduling, cross-modal fusion, and lightweight high-precision enhancement modules, the problem of the imbalance between real-time performance and accuracy in complex dynamic scenes of computer vision technology is solved. This enables efficient image enhancement in scenarios such as autonomous driving in vehicles, adapting to the needs of downstream tasks.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XUZHOU COLLEGE OF INDAL TECH
- Filing Date
- 2026-03-05
- Publication Date
- 2026-07-17
AI Technical Summary
Existing computer vision technologies struggle to balance real-time performance and enhanced accuracy in complex and dynamic scenarios. Traditional algorithms have limited applicability, deep learning models suffer from an imbalance between real-time performance and accuracy, lightweight technologies often sacrifice detail restoration, and cross-modal fusion technologies fail to effectively synergize with scene perception and lightweight enhancement. As a result, single technologies or simple overlay solutions cannot meet the full-process enhancement needs in complex scenarios.
A closed-loop system is constructed, consisting of a scene perception module, an enhancement scheduling module, a cross-modal fusion enhancement module, and a lightweight high-precision enhancement module. Through multi-feature fusion, dynamic scheduling, and hierarchical enhancement, it achieves full coverage adaptation for both extreme and normal scenarios. By combining attention mechanisms and lightweight convolutional structures, feature extraction and enhancement strategies are optimized to ensure both real-time performance and accuracy.
It achieves real-time image enhancement in complex and dynamic scenes, improves the accuracy of downstream tasks such as target detection and path planning, expands the application scope of the system, adapts to complex scenarios such as in-vehicle autonomous driving, and has high feasibility and application value.
Smart Images
Figure CN122415359A_ABST