Digital human image correction method, electronic equipment and storage medium
Through the digital human image processing method of multimodal feature analysis and adaptive correction strategy, the limitations of digital human image acquisition equipment, lighting sensitivity and insufficient algorithm processing capabilities are solved, and the digital human image quality improvement of high fidelity and reality is achieved, adapting to a variety of lighting conditions and scenes, and meeting the needs of film and television rendering and real-time interaction.
Patent Information
- Application Number
- CN202510353886.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-25
- Publication Date
- 2025-07-22
AI Technical Summary
The prior art has limitations in digital human image processing, lighting sensitivity, insufficient algorithm processing capabilities, and contradictions between real-time and generalization in digital human image processing, making it difficult to achieve high fidelity and realism in digital human image quality improvement.
The multimodal feature analysis and adaptive correction methods generated by image preprocessing, feature extraction, bias analysis and correction strategies are used to combine deep learning and traditional algorithms, and through spatial domain filtering, affine transformation, deep learning models and weighted fusion algorithms, the geometric deformation, color and texture of digital human images are optimized to build an adaptive correction strategy.
It has achieved the improvement of high fidelity and reality of digital human images, adapted to a variety of lighting conditions and scenes, met the needs of film and television rendering and real-time interaction, and improved the image quality by more than 20%.