Name
Abstract | ||
|---|---|---|
Various feature-based object recognition methods make use of similarity measures of features to guide the recognition process. These similarity measures often are only local in nature, meaning that the measures are derived from the local attributes of the features. A similarity measure is presented that takes the form of an object based on the position of the features. A quantity that assesses the similarity of features according to their position among all others, called a context similarity measure, is derived. It is tolerant to missing features or variations in their position. The primary interest is in measuring the similarity between model features and features extracted from an image. The authors consider the use of these measures for object recognition and, as an example, describe their application in a feature-based Hough transform. They show that the combination of local and context similarities considerably improves the recognition performance |
Year | DOI | Venue |
|---|---|---|
1993 | 10.1109/ICCV.1993.378204 | Berlin |
Keywords | Field | DocType |
Hough transforms,computer vision,feature extraction,object recognition,Hough transform,context similarity measure,contextual feature similarities,feature-based object recognition,model-based object recognition,similarity measures | Computer vision,3D single-object recognition,Similarity measure,Pattern recognition,Computer science,Hough transform,Feature extraction,Haar-like features,Context model,Feature (machine learning),Artificial intelligence,Cognitive neuroscience of visual object recognition | Conference |
Volume | Issue | ISBN |
1993 | 1 | 0-8186-3870-2 |
Citations | PageRank | References |
1 | 0.60 | 6 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Detlev Noll | 1 | 1 | 0.60 |
| Michael Schwarzinger | 2 | 1 | 0.60 |
| W von Seelen | 3 | 503 | 140.13 |