Image Feature Data Generation Using Intensity-Based Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for extracting feature points from images result in an excessively large number of feature points, leading to increased data processing time and complexity, especially when dealing with large image datasets and variations in scaling, rotation, and observation angles.
Innovation Solution
An apparatus and method that includes a feature point determination unit, a filtering unit to select significant feature points based on intensity and orientation, and a data generation unit to create binary feature vectors, reducing the amount of data and processing time while maintaining robustness to image transformations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional feature point extraction methods (LoG, DoG, Hessian matrix) are used, then feature points can be extracted from images, but an excessively large number of feature points are obtained, causing excessive data processing time and complexity
Solution Approach 1:
The patent changes the parameter of feature point selection by introducing a filtering mechanism that evaluates feature points based on intensity values and spatial distribution. The filtering unit selectively retains only significant feature points while discarding redundant ones, thereby reducing the total number of feature points processed without compromising the quality of extracted features.
2Reliability
If thousands of feature points are extracted per image, then comprehensive image features are captured, but the amount of feature data and processing time excessively increase
Solution Approach 1:
The patent extracts only the essential and significant feature points from the large set of initially detected feature points. The filtering unit identifies and extracts feature points that meet specific criteria (intensity thresholds, spatial distribution), separating them from redundant feature points. This selective extraction maintains reliable image feature representation while significantly reducing the data volume requiring further processing.
3Adaptability or versatility
If conventional feature data formation methods (SIFT, SURF) are used, then robust feature vectors are generated, but the operation process becomes complicated and requires excessively long computation time
Solution Approach 1:
The patent applies local quality by generating feature descriptors that are adapted to the specific characteristics of each filtered feature point. Instead of applying complex uniform processing to all feature points, the system creates localized descriptors based on the intensity and spatial properties of individual feature points, simplifying the overall operation process while maintaining robustness to image transformations.
Data Source
AI summary
The present invention relates to an apparatus and method for efficiently generating feature data which properly determines a feature point indicating features of images and describes the feature point. The apparatus for generating image feature data comprises: a feature point determining unit which determines a feature point from an image and extracts information on the determined feature point; a feature point filtering unit which determines, as a final feature point, at least one feature point from among the feature points determined by the feature point determining unit; and a feature data generating unit which generates image feature data based on the final feature points determined by the feature point filtering unit and feature point information on the final feature points.