2013
DOI: 10.1016/j.image.2013.06.005
|View full text |Cite
|
Sign up to set email alerts
|

Statistical texture retrieval in noise using complex wavelets

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
6
0

Year Published

2015
2015
2020
2020

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 12 publications
(6 citation statements)
references
References 35 publications
0
6
0
Order By: Relevance
“…In this work, the rotation and scale invariant properties are also validated for 12 directions and 5 scales in standard texture image database. Rakvongthai et al [17] have computed texture feature descriptor in the noisy environment based on the complex wavelets using the statistical values where each sub-band coefficients are modeled by the standard statistical distribution. To improve the retrieval accuracy, they have combined the magnitude and phase information of complex sub-band coefficients.…”
Section: B Texture Visual Feature Descriptorsmentioning
confidence: 99%
“…In this work, the rotation and scale invariant properties are also validated for 12 directions and 5 scales in standard texture image database. Rakvongthai et al [17] have computed texture feature descriptor in the noisy environment based on the complex wavelets using the statistical values where each sub-band coefficients are modeled by the standard statistical distribution. To improve the retrieval accuracy, they have combined the magnitude and phase information of complex sub-band coefficients.…”
Section: B Texture Visual Feature Descriptorsmentioning
confidence: 99%
“…Although our main contribution concerns magnitudes of complex wavelet coefficients, another experiment on influence of the information from angles has been conducted on the three datasets. To capture this information, we use von Mises distribution to characterize histograms of the relative phase as in [22], [23], and [25]. This joint modeling requires some changes in the similarity measurement.…”
Section: E Influence Of Phase Informationmentioning
confidence: 99%
“…Since magnitudes are positive values, it is obvious that all of this positive distributions are more suitable than a GGD to take advantage of the exponential family distributions and model the marginal behavior of complex wavelet coefficients histograms as well. Additionally, Weibull distribution has been successfully used to model magnitudes of complex wavelet coefficients in some previous texture retrieval models [22], [24], and [26]- [28]. Angles of complex coefficients are generally exploited using circular distributions such as Wrapped Cauchy and von Mises to characterize the relative phase [29], [30].…”
Section: Introductionmentioning
confidence: 99%
“…Each subband is composed by different complex numbers and it is shift invariant. We study in this paper the relative phases of these complex numbers [18]. From each complex numbers z = ai + b, the phase φ can be calculated as follows:…”
Section: A Complex Wavelet Decompositionmentioning
confidence: 99%