A Test-Bed for Region-Based Image Retrieval Using Multiple Segmentation Algorithms and the MPEG-7 eXperimentation Model: The Schema Reference System

Abstract

Abstract. The aim of the SCHEMA Network of Excellence is to bring together a critical mass of universities, research centers, industrial partners and end users, in order to design a reference system for contentbased semantic scene analysis, interpretation and understanding. In this paper, recent advances in the development of the SCHEMA reference system are reported, focusing on the application of region-based image retrieval using automatic segmentation. More specifically, the first and the second version of the reference system are presented and the motivation behind the different approaches followed during the development of these two versions is discussed. Experimental results for both systems, using a common collection of images, are shown. Additionally, a comparative evaluation of the two versions both in terms of retrieval accuracy and in terms of time-efficiency is performed, allowing the evaluation of the system as a whole as well as the evaluation of the usability of different components integrated with the reference system, such as the MPEG-7 eXperimentation Model. This illustrates the suitability of the SCHEMA reference system in serving as a test-bed for evaluating and comparing different algorithms and approaches pertaining to the content-based and semantic manipulation of visual information, ranging from segmentation algorithms to indexing features and methodologies.

Combining self-organizing neural nets with multivariate statistics for efficient color image retrieval

“…The utilized dataset was formed by pre-assigning the images into 20 distinct classes (e.g., cars, eagles, flowers, etc. ), as introduced in the Schema Reference System [26]. Following the common practice for measuring the performance in image retrieval systems, a subset of 50 query images was selected at random from this heterogeneous set, which included 1000 color images of 24 bpp each, given in portable pixel map format of sizes [192 · 128] or [128 · 192] pixels.…”

On the Perceptual Organization of Image Databases Using Cognitive Discriminative Biplots

“…Originally, it was formed by preassigning the images into C = 15 distinct classes (e.g., cars, eagles, flowers, etc.) as introduced in the SCHEMA reference system [25]. These are presented in Table 1.…”

A generic scheme for color image retrieval based on the multivariate Wald-Wolfowitz test

“…The utilized data set was formed by preassigning the images into 20 distinct classes (e.g., cars, eagles, flowers, etc. ), as introduced in the Schema Reference System [7]. By incorporating the coherent opinion of many individuals, 53 perceptually similar images were kept for each category.…”

“…These systems make use of low-level features such as color, shape, and texture to represent the image content. In addition, region-based retrieval systems such as NETRA toolbox [4], Blobworld system [5], ImageMap [6], Schema project [7], and WALRUS [8], make use of subimage region extraction and corresponding features to locate relevant images. Among others, MPEG-7 standard [9] defines a set of visual descriptors for image content representation designed to meet the requirements of different application domains.…”