Selection of Visual Descriptors for the Purpose of Multi-camera Object Re-identification - Feature Selection for Data and Pattern Recognition - page 262

Table 12.1 Characteristics of the visual feature descriptors
Feature name

Color space/representation Illumination inv. Geometric inv.

RGB Type 2 Ye s

MomentCentNorm

RGB Type 1, 2, 3, 4, 5 Ye s

MomentInvGPSO

VertTrace Transformed colour Type 1, 2, 3, 4, 5 No

RGB Type 1, 2 No

CLD

Transformed colour Type 1, 2, 3, 4, 5 No

CLDTrans

EHD Gray-level Type 1, 2, 3 No

RGB Type 1, 2, 3 No

LBPHist

rg-chromatic space Type 1 Ye s

Hist

HistFull RGB - Ye s

Gray-level Type 1, 2, 3, 4, 5 Ye s

Sift

Gray-level Type 1, 2, 3, 4, 5 Ye s

Surf64

Surf128 Gray-level Type 1, 2, 3, 4, 5 Ye s

Opponent colour space Type 1, 2, 3 Ye s

OpponentSift

Opponent colour space Type 1, 2, 3 Ye s

OpponentSurf64

CMSP HSV Type 1, 2, 3 No

0.2 0.2

0.15 0.15

0.1 0.1

0.05 0.05

0 0

0 50 200 150 200 250 0 50 200 150 200 250

0.2 0.2

0.15 0.15

0.1 0.1

0.05 0.05

0 0

0 50 100 150 200 250 0 50 100 150 200 250

0.2 0.2

0.15 0.15

0.1 0.1

0.05 0.05

0 0

0 50 100 150 200 250 0 50 100 150 200 250

Fig. 12.3 Colour histograms for example objects

This means that the values of R c and G c represent the share of red and green in
the original colour. It is easy to notice that deriving the share of blue is unnecessary,
as the three shares sum to one.