Face Feature Point Detection Using Segmented Eigen-Codebooks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional face feature point detection technologies are inefficient due to the need for individual extraction of local patterns, high computational complexity, and sensitivity to image brightness and illumination changes, leading to prolonged processing times and storage requirements.
Innovation Solution
A system comprising a training unit that generates an eigen-codebook representing correlations between feature vectors and feature points in facial images, and a detection unit that uses this eigen-codebook to rapidly and accurately detect feature points in target images by dividing images into segments, applying PCA, and calculating direction vectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If local patterns are individually extracted for each facial feature point, then feature point detection accuracy is improved, but processing time increases significantly and storage space is wasted
Solution Approach 1:
The patent divides facial images into multiple segments, with each segment corresponding to a specific facial feature point location. Instead of processing entire images or extracting complex local patterns, the system segments the image into manageable regions and processes each segment independently to detect feature points, thereby reducing processing time while maintaining detection accuracy.
Solution Approach 2:
The patent extracts only the necessary feature information from each image segment rather than processing entire images or storing complete local patterns. By extracting minimal sufficient features from segmented regions, the system reduces computational complexity and storage requirements while preserving the ability to accurately detect feature points.
2Measurement precision
If separate pattern searching algorithms are used for feature point detection, then detection accuracy is maintained, but computational complexity increases
Solution Approach 1:
The patent changes the parameters used for feature point detection by utilizing pre-trained eigen-codebooks and segment-based feature extraction rather than employing complex pattern searching algorithms. This parameter change simplifies the computational process while maintaining detection accuracy through the use of prepared reference data and simplified comparison operations.
3Ease of manufacture
If pixel value comparison methods are used for pattern search, then implementation is simple, but detection accuracy decreases under varying brightness and illumination conditions
Solution Approach 1:
The patent performs preliminary processing by pre-training eigen-codebooks using training images before actual detection. This preliminary action creates robust reference models that are insensitive to brightness and illumination variations, allowing the system to maintain high detection accuracy without relying on simple pixel value comparisons that are sensitive to lighting conditions.
Data Source
AI summary
Disclosed are a system and a method for detecting feature points of a face. The system for detecting feature points of a face include: a training unit configured to divide each of a plurality of facial images into segments of a predetermined size, and generate an eigen-codebook representing a correlation between a feature vector of each of the divided segments and location of a predetermined feature point in the facial image; and a detection unit configured to detect a facial area from a target image, divide the facial area into segments of a predetermined size, detect a feature vector of each of the divided segments of the facial area, and detect a location of a feature point in the facial area by using the detected feature vector and the eigen-codebook.