Facial Image Reference Matching for Blur and Distortion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing intelligent terminal devices struggle to optimize facial images with defects such as blur, overexposure, excessive darkness, or face distortion, as current image processing software is inadequate for these issues.
Innovation Solution
A method and apparatus for facial image processing that acquires a facial image to be optimized, retrieves a reference image from a preset face database, and optimizes the image based on the reference image, utilizing features like image resolution, brightness, facial expression, and facial contour to correct defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If professional image processing software is used to optimize defective facial images, then image quality can be improved, but device complexity and operation difficulty increase
Solution Approach 1:
The patent uses a pre-established three-dimensional facial model as a template or copy to represent the user's face. This 3D model serves as a reference that contains perfect facial geometry and features, allowing the system to reconstruct and optimize defective two-dimensional images by comparing them against the ideal 3D template, thereby improving image quality without requiring complex professional software
Solution Approach 2:
The patent replaces traditional 2D image processing algorithms with a 3D model-based approach. Instead of using complex image editing software that requires manual intervention, the system uses computational geometry and 3D rendering to automatically reconstruct the facial image, substituting mechanical image processing operations with 3D spatial reasoning and rendering operations
2Productivity
If existing image processing software is used for facial images with defects like blur or overexposure, then processing speed is maintained, but optimization effectiveness deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-establishing the user's three-dimensional facial model before image optimization is needed. This 3D model is created in advance and stored, so when a defective image needs optimization, the system can immediately retrieve the pre-computed 3D facial model and use it for rapid reconstruction, avoiding the need for time-consuming analysis during the optimization process itself
Solution Approach 2:
The patent introduces a three-dimensional facial model as an intermediary between the defective two-dimensional image and the optimal image representation. This 3D model acts as a mediator that contains the ground truth facial geometry, allowing the system to bridge the gap between the poor quality input image and the desired high-quality output without requiring complex iterative optimization algorithms
3Measurement precision
If reference images are retrieved from a database, then optimization accuracy is improved, but system complexity increases
Solution Approach 1:
The system performs self-service by automatically retrieving and selecting the appropriate three-dimensional facial model from the database based on the input image characteristics. The process is fully automated - the system identifies the user, retrieves their pre-computed 3D model, and uses it for optimization without requiring manual selection or intervention, thereby improving accuracy while keeping the user-side complexity low
Data Source
AI summary
Provided are a method and an apparatus for facial image processing, a computer device and a medium. The method for facial image processing includes acquiring a facial image to be optimized of a target user, acquiring a reference facial image matched with the facial image to be optimized from a preset face database, and optimizing the facial image to be optimized according to the reference facial image.