发明名称 |
Contextual weighting and efficient re-ranking for vocabulary tree based image retrieval |
摘要 |
Systems and methods are disclosed to search for a query image, by detecting local invariant features and local descriptors; retrieving best matching images by quantizing the local descriptors with a vocabulary tree; and reordering retrieved images with results from the vocabulary tree quantization. |
申请公布号 |
US8892542(B2) |
申请公布日期 |
2014.11.18 |
申请号 |
US201113339227 |
申请日期 |
2011.12.28 |
申请人 |
NEC Laboratories America, Inc. |
发明人 |
Wang Xiaoyu;Yang Ming;Cour Timothee;Zhu Shenghuo;Yu Kai |
分类号 |
G06F17/30;G06F7/00 |
主分类号 |
G06F17/30 |
代理机构 |
|
代理人 |
Kolodka Joseph |
主权项 |
1. A method to search for a query image, comprising
detecting local invariant features and local descriptors; retrieving best matching images by incorporating one or more contexts in matching quantized local descriptors with a vocabulary tree; reordering retrieved images with results from the vocabulary tree quantization; measuring similarity of local descriptors through image specific descriptor weighting; and comprising v, where weight wiq is defined based on the node counts along the quantization path of xi, comprising determiningwiq=∑v∈T(xi)ω(v)∑v∈T(xi)ω(v)×nq(v), where ω(v) is a weighting coefficient set to idƒ(v) empirically, where weight wiq depends on the descriptor only, and is shared for all nodes v along the path T(xi) and where nq(v) represents a number of descriptors in image q that are quantized to v. |
地址 |
Princeton NJ US |