Facial Composite Inference Using SVM Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low-resolution facial images captured by surveillance cameras pose challenges for accurate face recognition and identification due to limitations in resolution and lighting effects, making it difficult to generate a reliable facial composite for criminal identification.
Innovation Solution
A method and apparatus that utilize a support vector machine (SVM) to search for a facial type matching the extracted facial feature information from a facial composite database, generating a face shape model based on the searched type to serve as an initial facial composite model for criminal identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If super resolution technique is applied to correct low-resolution images, then image resolution is improved, but accurate face identification still cannot be achieved due to insufficient feature information
Solution Approach 1:
The system performs preliminary classification of facial types using SVM before attempting identification. By pre-categorizing faces into distinct types based on training data, the system prepares reference models in advance that can guide the identification process even when input image quality is poor.
Solution Approach 2:
The patent introduces facial type classification as an intermediary step between low-resolution image input and final face identification. This intermediate classification layer provides structural guidance that bridges the gap between limited image data and reliable identification results.
2Productivity
If facial feature extraction is performed on low-resolution images, then processing speed is improved, but identification reliability deteriorates due to loss of detailed feature information
Solution Approach 1:
The system segments the face identification process into two distinct stages: facial type classification (using simplified features for speed) and subsequent identification (using more detailed features). This segmentation allows each stage to operate at appropriate complexity levels, balancing speed and reliability.
Solution Approach 2:
The patent changes the parameter of feature extraction complexity based on the processing stage. In the classification stage, it uses coarser features for faster processing, while in the identification stage, it employs more detailed features to ensure reliability, thus adapting feature parameters to match processing requirements.
3Reliability
If a comprehensive facial composite database is created to improve identification accuracy, then system complexity increases due to large data storage and processing requirements
Solution Approach 1:
The patent segments the comprehensive facial database into organized categories of facial types. Rather than treating all faces uniformly, the system divides them into distinct typological groups, reducing the search space and simplifying the matching process while maintaining comprehensive coverage for accurate identification.
Data Source
AI summary
Provided is a method and apparatus for inferring a facial composite, whereby user's designation regarding at least one point of a facial image is received, facial feature information are extracted based on the received user's designation, a facial type that coincides with the extracted facial feature information is searched from a facial composite database to generate a face shape model based on the searched facial type and an initial facial composite model having a facial type similar to a face of the facial image from a low-resolution facial image through which face recognition or identification cannot be accurately performed is provided, so that the face shape model contributes to criminal arrest and a low-resolution facial image captured by a surveillance camera can be more efficiently used.