2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Workshops
DOI: 10.1109/cvpr.2005.433
|View full text |Cite
|
Sign up to set email alerts
|

Combining Local and Global Image Features for Object Class Recognition

Abstract: Object recognition is a central problem in computer vision

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
64
0
5

Publication Types

Select...
5
3
1

Relationship

0
9

Authors

Journals

citations
Cited by 112 publications
(69 citation statements)
references
References 11 publications
0
64
0
5
Order By: Relevance
“…В данной работе ВП сроится не по всему изображению в целом, а по особым точкам на изображении (далее -ОТ) [2,4]. ОТ на изображении считается точка (пиксель) с характерной окрестностью -т.е.…”
Section: признаковое описание изображенийunclassified
“…В данной работе ВП сроится не по всему изображению в целом, а по особым точкам на изображении (далее -ОТ) [2,4]. ОТ на изображении считается точка (пиксель) с характерной окрестностью -т.е.…”
Section: признаковое описание изображенийunclassified
“…Global features consist of the statistics extracted from a whole image or a bounding box, such as contour, shape, texture, colour or a combination of them [23]. They generally represent an image with a single high-dimensional feature vector and, thus, can be easily applied with any standard classification methods [24]. Moreover, thanks to the compact representation and low computational cost, they have been employed by some semantic mapping systems in real time.…”
Section: Global Featuresmentioning
confidence: 99%
“…Due to the different characteristics of local and global features, it is beneficial for some applications to combine both approaches. Lisin et al (Lisin et al, 2005) show two methods where combining local and global features improve the accuracy of a classification task. More than another hybrid-like approach has been found in the detection of regional features, in which regions are defined as arbitrary subsets of the image.…”
Section: Regional Symmetry Featuresmentioning
confidence: 99%