Facial Recognition Using Weighted Block Division
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current facial recognition systems face challenges in precision due to variations in facial images such as head tilt, illumination, and expression, often requiring computationally expensive methods that slow down the processing time and reduce efficiency.
Innovation Solution
A method that identifies statistical properties of pixels in a block of the subject image, converts them into feature vectors, and applies dimension compression to improve accuracy and efficiency by optimizing block division and weighting techniques, allowing for faster processing of high volumes of images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex normalizing algorithms or pixel-by-pixel analysis are applied to improve precision, then measurement precision improves, but productivity deteriorates due to computational expense and slow processing
Solution Approach 1:
The facial image is divided into multiple blocks, and statistical properties are calculated for each block independently. This segmentation allows the system to process images more efficiently by working with smaller units while maintaining recognition precision through aggregated block information.
Solution Approach 2:
The patent transforms the approach from analyzing individual pixel values to calculating statistical properties (mean, standard deviation, skewness, kurtosis) of pixel intensities within blocks. This parameter transformation reduces computational complexity while preserving essential facial features for recognition.
2Reliability
If traditional facial recognition methods are used to ensure accuracy, then reliability improves, but loss of time increases due to extensive computational requirements
Solution Approach 1:
The system pre-calculates and stores statistical properties of image blocks during an enrollment phase. When a recognition query is made, the system compares these pre-computed statistical features against the query image, significantly reducing real-time processing time while maintaining reliable identification accuracy.
Solution Approach 2:
The patent replaces traditional mechanical pixel-by-pixel comparison methods with a statistical approach using histograms and moment calculations. This substitution transforms the recognition process into mathematical operations that are computationally more efficient while maintaining or improving reliability.
Data Source
AI summary
A method, apparatus and computer program product for an improved facial recognition system are provided. Some embodiments may utilize a weighted block division of an image and capture a property measurement for pixels residing within a block. The measurements may be converted to vectors, compressed, and compared against compressed vectors of enrolled images to identify a characteristic or an image of a matching subject. Training processes may be utilized in order to optimize block divisions and weights.