Retrieval In Large Image Databases Research Articles

Effect of image partitioning on content-based image retrieval using colour and texture

The current application of content-based image retrieval (CBIR) was urgently needed to process image retrieval in large image databases. Previous research had succeeded in combining feature extraction methods using texture and colour. Combining the two methods of Gray Level Co-Occurrence matrix with Colour moment produces a higher level of precision than that produced when method is applied individually. This research examined more deeply about the effect of image partitioning using the grid partitioning method. The results of precision and recall on the CBIR method would be compared with no partitions, two partitions and three image partitions. The results obtained indicated that the addition of image partitions did not affect the precision value. Thus, it can be concluded that increasing the image partition increases the recall value of the overall CBIR performance.

Towards an interactive index structuring system for content-based image retrieval in large image databases

In recent years, the expansion of acquisition devices such as digital cameras, the development of storage and transmission techniques and the success of tablet computers facilitate the development of many large image databases as well as the interactions with the users. This thesis [1] deals with the problem of Content-Based Image Retrieval (CBIR) on these huge masses of data. Traditional CBIR systems generally rely on three phases: feature extraction, feature space structuring and retrieval. In this thesis, we are particularly interested in the structuring phase (normally called indexing phase), which plays a very important role in finding information in large databases. This phase aims at organizing the visual feature descriptors of all images into an efficient data structure in order to facilitate, accelerate and improve further retrieval. We assume that the feature extraction phase is completed and the image feature descriptors which are usually low-level features describing the color, shape, texture, etc. of all images are available. Instead of traditional structuring methods, clustering methods which organize image descriptors into groups of similar objects (clusters), without any constraint on the cluster size, are studied. The aim is to obtain an indexed structure more adapted to the retrieval of high dimensional and unbalanced data. Clustering process can be done without prior knowledge (unsupervised clustering) or with a limited amount of prior knowledge (semi-supervised clustering).

A Semi-Supervised Algorithm for Auto-Annotation and Unknown Structures Discovery in Satellite Image Databases

The increasing number and resolution of earth observation (EO) imaging sensors has had a significant impact on both the acquired image data volume and the information content in images. There is consequently a strong need for highly efficient search tools for EO image databases and for search methods to automatically identify and recognize structures within EO images. Content Based Image Retrieval (CBIR) and automatic image annotation systems have been designed to tackle the problem of image retrieval in large image databases. These two systems achieve a common goal, that is to learn the mapping function between low-level visual features and high-level image semantics. A setup, which has hardly been explored in annotating systems and which is the rule rather than the exception, is the case when the training database used to learn the mapping function is not exhaustive regarding semantic classes present in the images. This means that there exists unknown image classes for which there is no training examples in the training database. In this paper, we propose a semi-supervised method for auto-annotating satellite image databases and discovering unknown semantic image classes in these databases. The idea is to incorporate into the learning process the unannotated data which by definition contain the unknown image classes. The latter are considered to be latent structures in the data that appear when we train a hierarchical latent variable model with both the labeled and unlabeled data. We also show that, in our case, the use of unlabeled data leads to more reliable estimates regarding the model parameters. We present experimental results on a synthetic dataset, making a comparison of our algorithm with a semi-supervised Support Vector Machine (S3VM) on this dataset. We also demonstrate the effectiveness of our unknown image classes discovery procedure on a database of SPOT5 satellite images. We show that the results obtained on this database are rather positive since the new structures detected correspond to semantic classes which are not represented in the training database.

An efficient high-dimensional indexing method for content-based retrieval in large image databases

High-dimensional indexing methods have been proved quite useful for response time improvement. Based on Euclidian distance, many of them have been proposed for applications where data vectors are high-dimensional. However, these methods do not generally support efficiently similarity search when dealing with heterogeneous data vectors. In this paper, we propose a high-dimensional indexing method (KRA +-Blocks) as an extension of the region approximation approach to the kernel space. KRA +-Blocks combines nonlinear dimensionality reduction technique (KPCA) with region approximation approach to map data vectors into a reduced feature space. The created feature space is then used, on one hand to approximate regions, and on the other hand to provide an effective kernel distances for both filtering process and similarity measurement. In this way, the proposed approach achieves high performances in response time and in precision when dealing with high-dimensional and heterogeneous vectors.

FUZZY ART-BASED IMAGE CLUSTERING METHOD FOR CONTENT-BASED IMAGE RETRIEVAL

In this paper, an image clustering method that is essential for content-based image retrieval in large image databases efficiently is proposed by color, texture, and shape contents. The dominant triple HSV (Hue, Saturation, and Value), which are extracted from quantized HSV joint histogram in the image region, are used for representing color information in the image. Entropy and maximum entry from co-occurrence matrices are used for texture information and edge angle histogram is used for representing shape information. Due to its algorithmic simplicity and the several merits that facilitate the implementation of the neural network, Fuzzy ART has been exploited for image clustering. Original Fuzzy ART suffers unnecessary increase of the number of output neurons when the noise input is presented. Therefore, the improved Fuzzy ART algorithm is proposed to resolve the problem by differently updating the committed node and uncommitted node, and checking the vigilance test again. To show the validity of the proposed algorithm, experimental results on image clustering performance and comparison with original Fuzzy ART are presented in terms of recall rates.

Mental image search by boolean composition of region categories

Existing content-based image retrieval paradigms almost never address the problem of starting the search, when the user has no starting example image but rather a mental image. We propose a new image retrieval system to allow the user to perform mental image search by formulating boolean composition of region categories. The query interface is a region photometric thesaurus which can be viewed as a visual summary of salient regions available in the database. It is generated from the unsupervised clustering of regions with similar visual content into categories. In this thesaurus, the user simply selects the types of regions which should and should not be present in the mental image (boolean composition). The natural use of inverted tables on the region category labels enables powerful boolean search and very fast retrieval in large image databases. The process of query and search of images relates to that of documents with Google. The indexing scheme is fully unsupervised and the query mode requires minimal user interaction (no example image to provide, no sketch to draw). We demonstrate the feasibility of such a framework to reach the user mental target image with two applications: a photo-agency scenario on Corel Photostock and a TV news scenario. Perspectives will be proposed for this simple and innovative framework, which should motivate further development in various research areas.

Automatic image annotation and retrieval using weighted feature selection

The development of technology generates huge amounts of non-textual information, such as images. An efficient image annotation and retrieval system is highly desired. Clustering algorithms make it possible to represent visual features of images with finite symbols. Based on this, many statistical models, which analyze correspondence between visual features and words and discover hidden semantics, have been published. These models improve the annotation and retrieval of large image databases. However, image data usually have a large number of dimensions. Traditional clustering algorithms assign equal weights to these dimensions, and become confounded in the process of dealing with these dimensions. In this paper, we propose weighted feature selection algorithm as a solution to this problem. For a given cluster, we determine relevant features based on histogram analysis and assign greater weight to relevant features as compared to less relevant features. We have implemented various different models to link visual tokens with keywords based on the clustering results of K-means algorithm with weighted feature selection and without feature selection, and evaluated performance using precision, recall and correspondence accuracy using benchmark dataset. The results show that weighted feature selection is better than traditional ones for automatic image annotation and retrieval.

Image indexing and retrieval using signature trees

Significant research has focused on determining efficient methodologies for effective and speedy retrieval in large image databases. Towards that goal, the first contribution of this paper is an image abstraction technique, called variable-bin allocation (VBA), based on signature bitstrings and a corresponding similarity metric. The signature provides a compact representation of an image based on its color content and yields better retrieval effectiveness than when using classical global color histograms (GCHs) and comparable to the one obtained when using color-coherence vectors (CCVs). More importantly however, the use of VBA signatures allows savings of 75% and 87.5% in storage overhead when compared to GCHs and CCVs, respectively. The second contribution is the improvement upon an access structure, the S-tree, exploring the concept of logical and physical pages and a specialized nearest-neighbor type of algorithm, in order to improve retrieval speed. We compared the S-tree performance when indexing the VBA signatures against the SR-tree indexing GCHs and CCVs, since SR-trees are arguably the most efficient access method for high-dimensional points. Our experimental results, using a large number of images and varying several parameters, have shown that the combination VBA/S-tree outperforms the GCH/SR-tree combination in terms of effectiveness, access speed and size (up to 45%, 25% and 70% respectively). Due to the very high-dimensionality of the CCVs their indexing, even using an efficient access structure, the SR-tree, did not seem to be a feasible alternative.

Bayesian selection of the neighbourhood order for Gauss–Markov texture models

Gauss–Markov random fields have been successfully used as texture models in a host of applications, ranging from synthesis, feature extraction, classification and segmentation to query by image content and information retrieval in large image databases. An issue that deserves special consideration is the selection of the neighbourhood order (model complexity), which should faithfully reflect the Markovianity of spatial interactions. Estimating the parameters for the wrong model will not capture the essential statistical properties of the texture in question: a lower order model will not be informative enough, while a higher order will clutter the description with superfluous information, fitting the noise rather than the data. We give a full Bayesian solution for estimating the model complexity, using an appropriate set of prior probabilities on the parameters. The closed-form decision criterion is derived by employing a Gaussian approximation of the posterior probability around the mode. The validity and benefits of this approach are demonstrated on two important problems arising in machine vision: texture replication and image classification.

Visual information retrieval system via content-based approach

In this paper a content-based image retrieval method that can search large image databases efficiently by color, texture, and shape content is proposed. Quantized RGB histograms and the dominant triple (hue, saturation, and value), which are extracted from quantized HSV joint histogram in the local image region, are used for representing global/local color information in the image. Entropy and maximum entry from co-occurrence matrices are used for texture information and edge angle histogram is used for representing shape information. Relevance feedback approach, which has coupled proposed features, is used for obtaining better retrieval accuracy. A new indexing method that supports fast retrieval in large image databases is also presented. Tree structures constructed by k-means algorithm, along with the idea of triangle inequality, eliminate candidate images for similarity calculation between query image and each database image. We find that the proposed method reduces calculation up to average 92.2 percent of the images from direct comparison.

An efficient color representation for image retrieval

A compact color descriptor and an efficient indexing method for this descriptor are presented. The target application is similarity retrieval in large image databases using color. Colors in a given region are clustered into a small number of representative colors. The feature descriptor consists of the representative colors and their percentages in the region. A similarity measure similar to the quadratic color histogram distance measure is defined for this descriptor. The representative colors can be indexed in the three-dimensional (3-D) color space thus avoiding the high-dimensional indexing problems associated with the traditional color histogram. For similarity retrieval, each representative color in the query image or region is used independently to find regions containing that color. The matches from all of the query colors are then combined to obtain the final retrievals. An efficient indexing scheme for fast retrieval is presented. Experimental results show that this compact descriptor is effective and compares favorably with the traditional color histogram in terms of overall computational complexity.

Robust coding schemes for indexing and retrieval from large face databases

This paper introduces two new coding schemes, probabilistic reasoning models (PRM) and enhanced FLD (Fisher linear discriminant) models (EFM), for indexing and retrieval of large image databases with applications to face recognition. The unifying theme of the new schemes is that of lowering the space dimension ("data compression") subject to increased fitness for the discrimination index.