Refined Gabor Feature Extraction for Image Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-dimensional raw Gabor features in image recognition systems pose challenges in efficient processing and representation, limiting the effectiveness of existing image recognition technologies such as face recognition.
Innovation Solution
The proposed solution involves an apparatus and method for obtaining refined Gabor features through a Gabor filter, region selection, sub-region logic, feature calculation, and integral image processing, which reduces feature dimensionality by calculating magnitude sums or differences using Haar-patterns, facilitating easier feature extraction and classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If raw Gabor features are used for image recognition, then comprehensive image information is captured, but computational complexity and processing time increase significantly
Solution Approach 1:
The image is divided into multiple regions of interest (ROIs) based on wave-front methods or local region analysis. Gabor features are extracted separately for each region rather than processing the entire image uniformly, reducing overall computational complexity while maintaining recognition accuracy through localized feature capture.
Solution Approach 2:
The patent extracts and selects only the most discriminative Gabor features from the high-dimensional feature space. By identifying and retaining only the most relevant features for recognition tasks, the system reduces feature dimensionality and processing complexity while preserving the essential information needed for accurate recognition.
2Measurement precision
If high-dimensional raw Gabor features are processed, then detailed image characteristics are preserved, but processing speed decreases
Solution Approach 1:
Different regions of the image are processed with different levels of detail and feature extraction intensity. Critical regions receive more detailed analysis while less important regions use simplified processing, optimizing the balance between precision and processing speed across the entire image.
Solution Approach 2:
The system applies Gabor filtering and feature extraction to only the necessary portions of the image rather than processing every pixel uniformly. By focusing computational resources on regions most relevant to recognition, the system achieves adequate precision with reduced processing time.
3Loss of information
If full Gabor feature extraction is performed, then complete multi-scale and multi-orientation information is obtained, but memory requirements and computational load increase
Solution Approach 1:
The patent extracts and retains only the most informative Gabor features from the complete feature set. By identifying and removing redundant or less discriminative features, the system maintains essential image information while significantly reducing the volume of data that needs to be stored and processed.
Solution Approach 2:
Less critical Gabor features are discarded or compressed, while the most important features are preserved for recognition tasks. This selective retention strategy reduces memory requirements and computational load while maintaining the information necessary for accurate image recognition.
Data Source
AI summary
Machine-readable media, methods, apparatus and system for obtaining and processing image features are described. In some embodiments, groups of training features derived from regions of training images may be trained to obtain a plurality of classifiers, each classifier corresponding to each group of training features. The plurality of classifiers may be used to classify groups of validation features derived from regions of validation images to obtain a plurality of weights, wherein each weight corresponds to each region of the validation images and indicates how important the each region of the validation images is. Then, a weight may be discarded from the plurality of weights based upon a certain criterion.