CODE: Coherence Based Decision Boundaries for Feature Correspondence
A key challenge in feature correspondence is the difficulty in differentiating true and false matches at a local descriptor level. This forces adoption of strict similarity thresholds that discard many true matches. However, if analyzed at a global level, false matches are usually randomly scattered...
Main Authors: | , , , , , , |
---|---|
Format: | Journal article |
Language: | English |
Published: |
Institute of Electrical and Electronics Engineers
2017
|
_version_ | 1797053567410896896 |
---|---|
author | Lin, W Wang, F Cheng, M Yeung, S Torr, P Do, M Lu, J |
author_facet | Lin, W Wang, F Cheng, M Yeung, S Torr, P Do, M Lu, J |
author_sort | Lin, W |
collection | OXFORD |
description | A key challenge in feature correspondence is the difficulty in differentiating true and false matches at a local descriptor level. This forces adoption of strict similarity thresholds that discard many true matches. However, if analyzed at a global level, false matches are usually randomly scattered while true matches tend to be coherent (clustered around a few dominant motions), thus creating a coherence based separability constraint. This paper proposes a non-linear regression technique that can discover such a coherence based separability constraint from highly noisy matches and embed it into a correspondence likelihood model. Once computed, the model can filter the entire set of nearest neighbor matches (which typically contains over 90 percent false matches) for true matches. We integrate our technique into a full feature correspondence system which reliably generates large numbers of good quality correspondences over wide baselines where previous techniques provide few or no matches. |
first_indexed | 2024-03-06T18:45:27Z |
format | Journal article |
id | oxford-uuid:0e5a62ab-fb69-472f-a1e1-49d49595db62 |
institution | University of Oxford |
language | English |
last_indexed | 2024-03-06T18:45:27Z |
publishDate | 2017 |
publisher | Institute of Electrical and Electronics Engineers |
record_format | dspace |
spelling | oxford-uuid:0e5a62ab-fb69-472f-a1e1-49d49595db622022-03-26T09:45:36ZCODE: Coherence Based Decision Boundaries for Feature CorrespondenceJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:0e5a62ab-fb69-472f-a1e1-49d49595db62EnglishSymplectic Elements at OxfordInstitute of Electrical and Electronics Engineers2017Lin, WWang, FCheng, MYeung, STorr, PDo, MLu, JA key challenge in feature correspondence is the difficulty in differentiating true and false matches at a local descriptor level. This forces adoption of strict similarity thresholds that discard many true matches. However, if analyzed at a global level, false matches are usually randomly scattered while true matches tend to be coherent (clustered around a few dominant motions), thus creating a coherence based separability constraint. This paper proposes a non-linear regression technique that can discover such a coherence based separability constraint from highly noisy matches and embed it into a correspondence likelihood model. Once computed, the model can filter the entire set of nearest neighbor matches (which typically contains over 90 percent false matches) for true matches. We integrate our technique into a full feature correspondence system which reliably generates large numbers of good quality correspondences over wide baselines where previous techniques provide few or no matches. |
spellingShingle | Lin, W Wang, F Cheng, M Yeung, S Torr, P Do, M Lu, J CODE: Coherence Based Decision Boundaries for Feature Correspondence |
title | CODE: Coherence Based Decision Boundaries for Feature Correspondence |
title_full | CODE: Coherence Based Decision Boundaries for Feature Correspondence |
title_fullStr | CODE: Coherence Based Decision Boundaries for Feature Correspondence |
title_full_unstemmed | CODE: Coherence Based Decision Boundaries for Feature Correspondence |
title_short | CODE: Coherence Based Decision Boundaries for Feature Correspondence |
title_sort | code coherence based decision boundaries for feature correspondence |
work_keys_str_mv | AT linw codecoherencebaseddecisionboundariesforfeaturecorrespondence AT wangf codecoherencebaseddecisionboundariesforfeaturecorrespondence AT chengm codecoherencebaseddecisionboundariesforfeaturecorrespondence AT yeungs codecoherencebaseddecisionboundariesforfeaturecorrespondence AT torrp codecoherencebaseddecisionboundariesforfeaturecorrespondence AT dom codecoherencebaseddecisionboundariesforfeaturecorrespondence AT luj codecoherencebaseddecisionboundariesforfeaturecorrespondence |