Document Detail

Geodesic Invariant Feature (GIF): A Local Descriptor in Depth.
MedLine Citation:
PMID:  25494504     Owner:  NLM     Status:  Publisher    
Different from the photometric images, depth images resolve the distance ambiguity of the scene, while the properties, such as weak texture, high noise and low resolution, may limit the representation ability of the well-developed descriptors which are elaborately designed for the photometric images. In this paper, a novel depth descriptor, geodesic invariant feature (GIF), is presented for representing the parts of the articulate objects in depth images. GIF is a multi-level feature representation framework which is proposed based on the nature of depth images. Low-level, geodesic gradient is introduced to obtain the invariance to the articulate motion, such as scale and rotation variation. Mid-level, superpixel clustering is applied to reduce depth image redundancy, resulting in faster processing speed and better robustness to noise. High-level, deep network is used to exploit the nonlinearity of the data, which further improves the classification accuracy. The proposed descriptor is capable of encoding the local structures in the depth data effectively and efficiently. Comparisons with the state-of-the-art methods reveal the superiority of the proposed method.
Yazhou Liu; Pongsak Lasang; Mel Siegel; Quansen Sun
Related Documents :
25281474 - Improving the excess kurtosis (g2) method for localizing epileptic sources in magnetoen...
18850504 - Magnetic resonance imaging of the spine at 3 tesla.
12387844 - Assessment of the technology for functional imaging in cancer.
16639164 - Evidence for white matter abnormalities in schizophrenia.
4050934 - Magnetic resonance imaging of optic gliomas.
25423634 - Real-time, computer-assisted quantification of plus disease in retinopathy of prematuri...
Publication Detail:
Type:  JOURNAL ARTICLE     Date:  2014-12-04
Journal Detail:
Title:  IEEE transactions on image processing : a publication of the IEEE Signal Processing Society     Volume:  -     ISSN:  1941-0042     ISO Abbreviation:  IEEE Trans Image Process     Publication Date:  2014 Dec 
Date Detail:
Created Date:  2014-12-10     Completed Date:  -     Revised Date:  2014-12-11    
Medline Journal Info:
Nlm Unique ID:  9886191     Medline TA:  IEEE Trans Image Process     Country:  -    
Other Details:
Languages:  ENG     Pagination:  -     Citation Subset:  -    
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine

Previous Document:  Random Forest Construction with Robust Semi-supervised Node Splitting.
Next Document:  Detection of Dynamic Background due to Swaying Motions from Motion Features.