View source for 20240029287. Matching Local Image Feature Descriptors in Image Analysis simplified abstract (Imagination Technologies Limited)

=Matching Local Image Feature Descriptors in Image Analysis=


==Organization Name==

[[:Category:Imagination Technologies Limited|Imagination Technologies Limited]]


[[Category:Imagination Technologies Limited]]

==Inventor(s)==

[[:Category:Ruan Lakemond of Cheltenham (AU)|Ruan Lakemond of Cheltenham (AU)]][[Category:Ruan Lakemond of Cheltenham (AU)]]

[[:Category:Timothy Smith of London (GB)|Timothy Smith of London (GB)]][[Category:Timothy Smith of London (GB)]]

==Matching Local Image Feature Descriptors in Image Analysis - A simplified explanation of the abstract==


This abstract first appeared for US patent application 20240029287 titled 'Matching Local Image Feature Descriptors in Image Analysis

==Simplified Explanation==

The abstract describes a method for matching features in images captured from different camera viewpoints using epipolar geometry. The method involves comparing local descriptors of features in the images to determine similarity measures. The method then identifies the best and next-best geometric matches to a feature in the first image, as well as a global best match. It performs comparisons of similarity measures between the geometric best match and the global best match, as well as between the geometric best match and the geometric next-best match. If certain thresholds are met, the method selects the geometric best match feature in the second image.

* Method uses epipolar geometry to define a geometrically-constrained region in a second image corresponding to a feature in a first image.
* Compares local descriptors of features in the second image with the local descriptor of the feature in the first image to determine similarity measures.
* Identifies the geometric best match and geometric next-best match to the feature in the second image, as well as the global best match.
* Performs comparisons of similarity measures between the geometric best match and the global best match, and between the geometric best match and the geometric next-best match.
* Selects the geometric best match feature in the second image if certain thresholds are met.

Potential applications of this technology:
* Object recognition and tracking in computer vision systems.
* Augmented reality applications.
* Image stitching and panorama creation.
* 3D reconstruction from multiple images.

Problems solved by this technology:
* Matching features across different camera viewpoints.
* Handling geometric constraints in feature matching.
* Improving accuracy and robustness of feature matching algorithms.

Benefits of this technology:
* Improved accuracy in feature matching.
* Robustness to changes in viewpoint and lighting conditions.
* Efficient and reliable matching of features in images.
* Enables various computer vision applications such as object recognition and 3D reconstruction.


==Original Abstract Submitted==

a method of feature matching in images captured from camera viewpoints uses the epipolar geometry of the viewpoints to define a geometrically-constrained region in a second image corresponding to a first feature in a first image; comparing the local descriptor of the first feature with local descriptors of features in the second image to determine respective measures of similarity; identifying, from the features located in the geometrically-constrained region, (i) a geometric best match and (ii) a geometric next-best match to the first feature; identifying a global best match to the first feature; performing a first comparison of the measures of similarity for the geometric best match and the global best match; performing a second comparison of the measures of similarity for the geometric best match and the geometric next-best match; and, if thresholds are met, selecting the geometric best match feature in the second image.



[[Category:G06T7/593]]
[[Category:G06T7/73]]
[[Category:G06V20/10]]
[[Category:G06F18/22]]