Image representation method and processing device based on local PCA whitening

Patent metadata

Field	Value
Publication number	US-10424052-B2
Application number	US-201515756193-A
Country	US
Kind code	B2
Filing date	Sep 15, 2015
Priority date	Sep 15, 2015
Publication date	Sep 24, 2019
Grant date	Sep 24, 2019

Abstract

An image representation method and processing device based on local PCA whitening. A first mapping module maps words and characteristics to a high-dimension space. A principal component analysis module conducts principal component analysis in each corresponding word space, to obtain a projection matrix. A VLAD computation module computes a VLAD image representation vector; a second mapping module maps the VLAD image representation vector to the high-dimension space. A projection transformation module conducts projection transformation on the VLAD image representation vector obtained by means of projection. A normalization processing module conducts normalization on characteristics obtained by means of projection transformation, to obtain a final image representation vector. An obtained image representation vector is projected to a high-dimension space first, then projection transformation is conducted on a projection matrix computed in advance and vectors corresponding to words, to obtain a low-dimension vector; and in this way, the vectors corresponding to the words are consistent. The disclosed method and the processing device can obtain better robustness and higher performance.

First claim

What is claimed is:

1. An image representation method based on regional PCA whitening, comprising:
constructing a vocabulary, assigning each feature to a corresponding word and mapping words and features to a high dimensional space, wherein dimensions of the high dimensional space are higher than dimensions of the current space of words and features;
conducting principal component analysis in each corresponding word space to obtain a projection matrix;
computing VLAD image representation vectors according to the vocabulary;
mapping the VLAD image representation vectors to the high dimensional space;
conducting projection transformation, according to the projection matrix, on VLAD image representation vectors obtained by means of projection; and
normalizing features acquired by means of projection transformation to obtain final image representation vectors.

2. The method according to claim 1 , wherein the vocabulary is constructed by K-means algorithm and each feature is assigned to its nearest words, wherein the feature obtained by means of projection transformation is performed with second normal form normalization to obtain final image representation vectors.

3. The method according to claim 1 , wherein the projection matrix may be obtained by conducting principal component analysis in each corresponding word space, which specifically includes:
computing a transition matrix G i with a formula below

G
i

=

1
D

⁢

∑

j
=
1

,

k
=
1

⁢

⁢

(

x
j

-

c
i

)

⁢

(

x
k

-

c
i

)

T

,

where c i is the i-the word, x is the features assigned to the word, D is feature dimensionality;
performing eigen-decomposition on the matrix G i with formulas below so as to obtained the eigenvalues eigval(G i ) and eigenvectors eigvect(G i ); and
(λ 1 i ,λ 2 i , . . . ,λ D i )=eigval( G i )

( u 1 i ,u 2 i , . . . ,u D i )=eigvect( G i )

computing the projection matrix P t i with a formula below,
P t i =L t i U t i

where

L
t
i

=

diag
⁡

(

1

λ
1
i

+
ϵ

,

1

λ
2
i

+
ϵ

,
…
⁢

,

1

λ
t
i

+
ϵ

)

,

⁢

U
t
i

=

[

u
1
i

,

u
2
i

,
…
⁢

,

u
t
i

]

,

ε and t are preset parameters.

4. The method according to claim 3 , wherein the VLAD image representation vectors are mapped to the high dimensional space with a formula below:
ψ κ ( x )= e iτ log x √{square root over ( x sech(πτ))},

where τ denotes the index of mapping.

5. The method according to claim 3 , wherein the step of conducting projection transformation, according to the projection matrix, on VLAD image representation vectors obtained by means of projection includes: conducting the projection transformation a formula below to obtain feature y,
y =[ P t 1 x 1 ,P t 2 x 2 , . . . ,P t k x k ].

6. An image representation processing device based on regional PCA whitening, comprising:
a first mapping module configured to construct a vocabulary, assigning each feature to a corresponding word and mapping words and features to a high dimensional space, wherein dimensions of the high dimensional space are higher than dimensions of the current space of words and features;
a PCA module configured to conduct principal component analysis in each corresponding word space to obtain a projection matrix;
a VLAD computation module configured to compute VLAD image representation vectors according to the vocabulary;
a s

Assignees

• Univ Peking Shenzhen Graduate School

Inventors

• Wang Wenmin
• Zhen Mingmin
• Wang Ronggang
• Li Ge
• Dong Shengfu
• Wang Zhenyu
• Li Ying
• Gao Wen

Classifications

• G06V10/7715 Primary

  Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods · CPC title

• G06F18/2135

  based on approximation criteria, e.g. principal component analysis · CPC title

• G06F18/24

  Classification techniques · CPC title

• G06V10/464

  using a plurality of salient features, e.g. bag-of-words [BoW] representations · CPC title

• G06T7/33

  using feature-based methods · CPC title