You are on page 1of 3

Computational Perceptual Features for Texture Representation and Retrieval

A perception-based approach to content-based image representation and retrieval is proposed in this paper. We consider textured images and propose to model their textural content by a set of features having a perceptual meaning and their application to contentbased image retrieval. We present a new method to estimate a set of perceptual textural features, namely coarseness, directionality, contrast, and busyness. The proposed computational measures can be based upon two representations: the original images representation and the autocorrelation function (associated with original images representation. The set of computational measures proposed is applied to content-based image retrieval on a large image data set, the well-!nown "rodat# database. $xperimental results and benchmar!ing show interesting performance of our approach. %irst, the correspondence of the proposed computational measures to human &udgments is shown using a psychometric method based upon the 'pearman ran!-correlation coefficient. 'econd, the application of the proposed computational measures in texture retrieval shows interesting results, especially when using results fusion returned by each of the two representations. (omparison is also given with related wor!s and show excellent performance of our approach compared to related approaches on both sides: correspondence of the proposed computational measures with human &udgments as well as the retrieval effectiveness.

Existing System:
)n earlier days, image retrieving from large image database can be done by following ways. We will discuss briefly about the image retrieving of various steps Automatic )mage Annotation and *etrieval using (ross +edia *elevance +odels (oncept "ased ,uery $xpansion

,uery 'ystem "ridging The 'emantic -ap %or .arge )mage /atabases 0ntology-"ased ,uery $xpansion Widget for information *etrieval /etecting image purpose in World-Wide Web documents

Proposed System: *elevance feedbac! is an interactive process that starts with normal (")*. The user input a 1uery, and then the system extracts the image feature and measure the distance with images in the database. An initial retrieval list is then generated. 2ser can choose the relevant image to further refine the 1uery, and this process can be iterated many times until the user find the desired images.

,uery update

2ser3s %eedbac!

)nput ,uery

%eature $xtraction

'imilarity measure

*etrieval *esult

%ind all images4

%inal *etrieval *esult

AL !R"T#$ %SE&:
T'e ()means algorit'm: Algorit'm: !-means. The !-means algorithm for partitioning based on the mean value of the ob&ects in the cluster. "nput: The number of clusters ! and a database containing n ob&ects.

!utput: A set of ! clusters that minimi#es the s1uared-error criterion. $et'od: (5 arbitrarily choose ! ob&ects as the initial cluster centers: (6 repeat (7 (re assign each ob&ect to the cluster to which the ob&ect is the most similar, based on the mean value of the ob&ects in the cluster8 (9 2pdate the cluster means, i.e., calculate the mean value of the ob&ects for each cluster8 (: 2ntil no change.

#ard*are Re+uirements:
; ; ; ; ; ; 'ystem >ard /is! %loppy /rive +onitor +ouse *am : <entium )= 6.9 ->#. : 9? -". : 5.99 +b. : 5: =-A (olour. : .ogitech. : 6:@ +b.

Soft*are Re+uirements:
; ; ; 0perating system (oding .anguage Tool 2sed : - Windows A< <rofessional. : - Bava. : - $clipse.

Coureddine Abbadeni, D(omputational <erceptual %eatures for Texture *epresentation and *etrievalE, "EEE Transactions on "mage Processing- .ol/01- ,o/2- 3anuary 0122/