Upgrading Color Distributions for Image Retrieval Can We Do Better?

Vertan, Constantin; Boujemaa, Nozha

doi:10.1007/3-540-40053-2_16

Cited by 13 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This experience is carried out to show the importance of considering the scaling factors of the Daubechies-8 decomposed versions of the histogramsh e a andh e b in the querying. Each human subject is asked to formulate a query from the database and to execute a querying using N = 5 feature histograms which are h L , h h a , h h b ,h e a andh e b to represent the query color image, while computing the metric weightsw l 0 and {w l k } 7 k=0 by the logistic regression for each l ∈ {1, 2, 3, 4, 5}, and to give a goodness score to each retrieved image, then to reformulate a query from the database and to execute a querying using the same feature histograms, while keeping the metric termsw 0 |Q 4 [0] −T 4 [0]| and w 0 |Q 5 [0] −T 5 [0]| equal to zero by affecting a zero tow 4 0 andw 5 0 , and to give a goodness score to each retrieved image. Each querying is repeated twenty times by choosing a new query from the database each time.…”

Section: Experimental Results and Evaluationmentioning

confidence: 99%

“…The resulted precision-scope curves for each compression order are: Thanks to the above precision-scope curves, we can notice the degradation of the querying when we neglect the scaling factors. In fact, because of the quantization if we keep the metric termsw 0 |Q 4 5 [0]| equal to zero by affecting a zero tow 4 0 andw 5 0 , the metric will not distinguish the two color component multispectral gradient module mean histograms Q 4 and T 4 or Q 5 and T 5 having the same curvature variations at the same X-coordinates, but these curvatures have different magnitudes. Consequently, the metric can not distinguish two LAB color images having almost similar colors and luminances, but different object shapes.…”

Section: Experimental Results and Evaluationmentioning

confidence: 99%

“…The multispectral gradient module mean histogram is inspired from the Laplacian weighted histogram which is proposed by Ref. 5. This latter is a good tool to distinguish the pixels located at the neighborhoods of color image edges.…”

Section: Weighted Histogramsmentioning

confidence: 99%

See 2 more Smart Citations

Image Retrieval Based on Region Separation and Multiresolution Analysis

Ksantini

Ziou

Dubeau

2006

Int. J. Wavelets Multiresolut Inf. Process.

View full text Add to dashboard Cite

In this paper, a simple and fast querying method for content-based image retrieval is represented. Using the multispectral gradient, a color image is split into two disjoint parts that are the homogeneous color regions and the edge regions. The homogeneous regions are represented by the traditional color histograms, and the edge regions are represented by the multispectral gradient module mean histograms. In order to measure the similarity degree between two color images both quickly and effectively, we use a one-dimensional pseudo-metric, which makes use of the one-dimensional Daubechies decomposition and compression of the extracted histograms. Our querying method is invariant to the query color image object translations and color intensities. The experimental results are reported on a collection of 10,000 LAB color images.

show abstract

Section: Experimental Results and Evaluationmentioning

confidence: 99%

Section: Experimental Results and Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

Image Retrieval Based on Region Separation and Multiresolution Analysis

Ksantini

Ziou

Dubeau

2006

Int. J. Wavelets Multiresolut Inf. Process.

View full text Add to dashboard Cite

show abstract

“…We use weighted histograms [33], where the contribution of each pixel is proportional to its importance in the local context. As weighting functions, we use the Laplacian kÁðx; yÞk 2 at the pixel ðx; yÞ to emphasize corners and edges, and the probability of the color of the current pixel in a local window, with a small value signaling importance.…”

Section: Global Descriptorsmentioning

confidence: 99%

A Statistical Framework for Image Category Search from a Mental Picture

Ferecatu

Geman

2009

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Abstract-Starting from a member of an image database designated the "query image," traditional image retrieval techniques, for example, search by visual similarity, allow one to locate additional instances of a target category residing in the database. However, in many cases, the query image or, more generally, the target category, resides only in the mind of the user as a set of subjective visual patterns, psychological impressions, or "mental pictures." Consequently, since image databases available today are often unstructured and lack reliable semantic annotations, it is often not obvious how to initiate a search session; this is the "page zero problem." We propose a new statistical framework based on relevance feedback to locate an instance of a semantic category in an unstructured image database with no semantic annotation. A search session is initiated from a random sample of images. At each retrieval round, the user is asked to select one image from among a set of displayed images-the one that is closest in his opinion to the target class. The matching is then "mental." Performance is measured by the number of iterations necessary to display an image which satisfies the user, at which point standard techniques can be employed to display other instances. Our core contribution is a Bayesian formulation which scales to large databases. The two key components are a response model which accounts for the user's subjective perception of similarity and a display algorithm which seeks to maximize the flow of information. Experiments with real users and two databases of 20,000 and 60,000 images demonstrate the efficiency of the search process.

show abstract

“…The Columbia set contains 100 categories of objects each one represented by 72 images. Each image is encoded using the laplacian color descriptor [22] resulting into a feature vector of 216 coefficients.…”

Section: B Database Categorizationmentioning

confidence: 99%

Validity of Fuzzy Clustering Using Entropy Regularization

Sahbi

Boujemaa

The 14th IEEE International Conference on Fuzzy Systems, 2005. FUZZ '05.

View full text Add to dashboard Cite

We introduce in this paper a new formulation of the regularized fuzzy C-means (FCM) algorithm which allows us to find automatically the actual number of clusters. The approach is based on the minimization of an objective function which mixes, via a particular parameter, a classical FCM term and a new entropy regularizer. The main contribution of the method is the introduction of a new exponential form of the fuzzy memberships which ensures the consistency of their bounds and makes it possible to interpret the mixing parameter as the variance (or scale) of the clusters. This variance closely related to the number of clusters, provides us with an intuitive and an easy to set parameter. We will discuss the proposed approach from the regularization point-of-view and we will demonstrate its validity both analytically and experimentally. We will show an extension of the method to non-linearly separable data. Finally, we will illustrate preliminary results both on simple toy examples as well as database categorization problems.

show abstract

Upgrading Color Distributions for Image Retrieval Can We Do Better?

Cited by 13 publications

References 12 publications

Image Retrieval Based on Region Separation and Multiresolution Analysis

Image Retrieval Based on Region Separation and Multiresolution Analysis

A Statistical Framework for Image Category Search from a Mental Picture

Validity of Fuzzy Clustering Using Entropy Regularization

Contact Info

Product

Resources

About