The mosaic test:benchmarking colour-based image retrieval systems using image mosaics

Evaluation and benchmarking in content-based image retrieval has always been a somewhat neglected research area, making it difficult to judge the efficacy of many presented approaches. In this paper we investigate the issue of benchmarking for colour-based image retrieval systems, which enable users to retrieve images from a database based on lowlevel colour content alone. We argue that current image retrieval evaluation methods are not suited to benchmarking colour-based image retrieval systems, due in main to not allowing users to reflect upon the suitability of retrieved images within the context of a creative project and their reliance on highly subjective ground-truths. As a solution to these issues, the research presented here introduces the Mosaic Test for evaluating colour-based image retrieval systems, in which test-users are asked to create an image mosaic of a predetermined target image, using the colour-based image retrieval system that is being evaluated. We report on our findings from a user study which suggests that the Mosaic Test overcomes the major drawbacks associated with existing image retrieval evaluation methods, by enabling users to reflect upon image selections and automatically measuring image relevance in a way that correlates with the perception of many human assessors. We therefore propose that the Mosaic Test be adopted as a standardised benchmark for evaluating and comparing colour-based image retrieval systems.

Exploring the effectiveness of similarity-based visualisations for colour-based image retrieval

In April 2009, Google Images added a filter for narrowing search results by colour. Several other systems for searching image databases by colour were also released around this time. These colour-based image retrieval systems enable users to search image databases either by selecting colours from a graphical palette (i.e., query-by-colour), by drawing a representation of the colour layout sought (i.e., query-by-sketch), or both. It was comments left by readers of online articles describing these colour-based image retrieval systems that provided us with the inspiration for this research. We were surprised to learn that the underlying query-based technology used in colour-based image retrieval systems today remains remarkably similar to that of systems developed nearly two decades ago. Discovering this ageing retrieval approach, as well as uncovering a large user demographic requiring image search by colour, made us eager to research more effective approaches for colour-based image retrieval. In this thesis, we detail two user studies designed to compare the effectiveness of systems adopting similarity-based visualisations, query-based approaches, or a combination of both, for colour-based image retrieval. In contrast to query-based approaches, similarity-based visualisations display and arrange database images so that images with similar content are located closer together on screen than images with dissimilar content. This removes the need for queries, as users can instead visually explore the database using interactive navigation tools to retrieve images from the database. As we found existing evaluation approaches to be unreliable, we describe how we assessed and compared systems adopting similarity-based visualisations, query-based approaches, or both, meaningfully and systematically using our Mosaic Test - a user-based evaluation approach in which evaluation study participants complete an image mosaic of a predetermined target image using the colour-based image retrieval system under evaluation.

Leveraging Multiple Features for Image Retrieval and Matching

The goal of image retrieval and matching is to find and locate object instances in images from a large-scale image database. While visual features are abundant, how to combine them to improve performance by individual features remains a challenging task. In this work, we focus on leveraging multiple features for accurate and efficient image retrieval and matching. We first propose two graph-based approaches to rerank initially retrieved images for generic image retrieval. In the graph, vertices are images while edges are similarities between image pairs. Our first approach employs a mixture Markov model based on a random walk model on multiple graphs to fuse graphs. We introduce a probabilistic model to compute the importance of each feature for graph fusion under a naive Bayesian formulation, which requires statistics of similarities from a manually labeled dataset containing irrelevant images. To reduce human labeling, we further propose a fully unsupervised reranking algorithm based on a submodular objective function that can be efficiently optimized by greedy algorithm. By maximizing an information gain term over the graph, our submodular function favors a subset of database images that are similar to query images and resemble each other. The function also exploits the rank relationships of images from multiple ranked lists obtained by different features. We then study a more well-defined application, person re-identification, where the database contains labeled images of human bodies captured by multiple cameras. Re-identifications from multiple cameras are regarded as related tasks to exploit shared information. We apply a novel multi-task learning algorithm using both low level features and attributes. A low rank attribute embedding is joint learned within the multi-task learning formulation to embed original binary attributes to a continuous attribute space, where incorrect and incomplete attributes are rectified and recovered. To locate objects in images, we design an object detector based on object proposals and deep convolutional neural networks (CNN) in view of the emergence of deep networks. We improve a Fast RCNN framework and investigate two new strategies to detect objects accurately and efficiently: scale-dependent pooling (SDP) and cascaded rejection classifiers (CRC). The SDP improves detection accuracy by exploiting appropriate convolutional features depending on the scale of input object proposals. The CRC effectively utilizes convolutional features and greatly eliminates negative proposals in a cascaded manner, while maintaining a high recall for true objects. The two strategies together improve the detection accuracy and reduce the computational cost.

Content Based Image Retrieval System for Malayalam Handwritten Characters

Content Based Image Retrieval is one of the prominent areas in Computer Vision and Image Processing. Recognition of handwritten characters has been a popular area of research for many years and still remains an open problem. The proposed system uses visual image queries for retrieving similar images from database of Malayalam handwritten characters. Local Binary Pattern (LBP) descriptors of the query images are extracted and those features are compared with the features of the images in database for retrieving desired characters. This system with local binary pattern gives excellent retrieval performance

Content-Based Image Retrieval of Axial Brain Slices Using a Novel LBP with a Ternary Encoding

Retrieval of similar anatomical structures of brain MR images across patients would help the expert in diagnosis of diseases. In this paper, modified local binary pattern with ternary encoding called modified local ternary pattern (MOD-LTP) is introduced, which is more discriminant and less sensitive to noise in near-uniform regions, to locate slices belonging to the same level from the brain MR image database. The ternary encoding depends on a threshold, which is a user-specified one or calculated locally, based on the variance of the pixel intensities in each window. The variancebased local threshold makes the MOD-LTP more robust to noise and global illumination changes. The retrieval performance is shown to improve by taking region-based moment features of MODLTP and iteratively reweighting the moment features of MOD-LTP based on the user’s feedback. The average rank obtained using iterated and weighted moment features of MOD-LTP with a local variance-based threshold, is one to two times better than rotational invariant LBP (Unay, D., Ekin, A. and Jasinschi, R.S. (2010) Local structure-based region-of-interest retrieval in brain MR images. IEEE Trans. Inf. Technol. Biomed., 14, 897–903.) in retrieving the first 10 relevant images

Image retrieval techniques for THz applications in cultural heritage

Techniques to retrieve reliable images from complicated objects are described, overcoming problems introduced by uneven surfaces, giving enhanced depth resolution and improving image contrast. The techniques are illustrated with application to THz imaging of concealed wall paintings.

Medical image retrieval based on complexity analysis

Texture is one of the most important visual attributes used in image analysis. It is used in many content-based image retrieval systems, where it allows the identification of a larger number of images from distinct origins. This paper presents a novel approach for image analysis and retrieval based on complexity analysis. The approach consists of a texture segmentation step, performed by complexity analysis through BoxCounting fractal dimension, followed by the estimation of complexity of each computed region by multiscale fractal dimension. Experiments have been performed with MRI database in both pattern recognition and image retrieval contexts. Results show the accuracy of the method and also indicate how the performance changes as the texture segmentation process is altered.

Image Retrieval: The MIRACLE Approach

ImageCLEF is a pilot experiment run at CLEF 2003 for cross language image retrieval using textual captions related to image contents. In this paper, we describe the participation of the MIRACLE research team (Multilingual Information RetrievAl at CLEF), detailing the different experiments and discussing their preliminary results.

A scalable approach for content based image retrieval in cloud datacenter

The emergence of cloud datacenters enhances the capability of online data storage. Since massive data is stored in datacenters, it is necessary to effectively locate and access interest data in such a distributed system. However, traditional search techniques only allow users to search images over exact-match keywords through a centralized index. These techniques cannot satisfy the requirements of content based image retrieval (CBIR). In this paper, we propose a scalable image retrieval framework which can efficiently support content similarity search and semantic search in the distributed environment. Its key idea is to integrate image feature vectors into distributed hash tables (DHTs) by exploiting the property of locality sensitive hashing (LSH). Thus, images with similar content are most likely gathered into the same node without the knowledge of any global information. For searching semantically close images, the relevance feedback is adopted in our system to overcome the gap between low-level features and high-level features. We show that our approach yields high recall rate with good load balance and only requires a few number of hops.

Automatic texture segmentation for texture-based image retrieval

Texture-segmentation is the crucial initial step for texture-based image retrieval. Texture is the main difficulty faced to a segmentation method. Many image segmentation algorithms either can’t handle texture properly or can’t obtain texture features directly during segmentation which can be used for retrieval purpose. This paper describes an automatic texture segmentation algorithm based on a set of features derived from wavelet domain, which are effective in texture description for retrieval purpose. Simulation results show that the proposed algorithm can efficiently capture the textured regions in arbitrary images, with the features of each region extracted as well. The features of each textured region can be directly used to index image database with applications as texture-based image retrieval.

The mosaic test:measuring the effectiveness of colour-based image retrieval

A variety of content-based image retrieval systems exist which enable users to perform image retrieval based on colour content - i.e., colour-based image retrieval. For the production of media for use in television and film, colour-based image retrieval is useful for retrieving specifically coloured animations, graphics or videos from large databases (by comparing user queries to the colour content of extracted key frames). It is also useful to graphic artists creating realistic computer-generated imagery (CGI). Unfortunately, current methods for evaluating colour-based image retrieval systems have 2 major drawbacks. Firstly, the relevance of images retrieved during the task cannot be measured reliably. Secondly, existing methods do not account for the creative design activity known as reflection-in-action. Consequently, the development and application of novel and potentially more effective colour-based image retrieval approaches, better supporting the large number of users creating media for use in television and film productions, is not possible as their efficacy cannot be reliably measured and compared to existing technologies. As a solution to the problem, this paper introduces the Mosaic Test. The Mosaic Test is a user-based evaluation approach in which participants complete an image mosaic of a predetermined target image, using the colour-based image retrieval system that is being evaluated. In this paper, we introduce the Mosaic Test and report on a user evaluation. The findings of the study reveal that the Mosaic Test overcomes the 2 major drawbacks associated with existing evaluation methods and does not require expert participants. © 2012 Springer Science+Business Media, LLC.

Content-Based Image Retrieval for Computer Tomography Images Using Wavelet Descriptors

An approach to building a CBIR-system for searching computer tomography images using the methods of wavelet-analysis is presented in this work. The index vectors are constructed on the basis of the local features of the image and on their positions. The purpose of the proposed system is to extract visually similar data from the individual personal records and from analogous analysis of other patients.

Image retrieval on the Honeycomb Image Browser

Efficient and effective approaches of dealing with the vast amount of visual information available nowadays are highly sought after. This is particularly the case for image collections, both personal and commercial. Due to the magnitude of these ever expanding image repositories, annotation of all images images is infeasible, and search in such an image collection therefore becomes inherently difficult. Although content-based image retrieval techniques have shown much potential, such approaches also suffer from various problems making it difficult to adopt them in practice. In this paper, we follow a different approach, namely that of browsing image databases for image retrieval. In our Honeycomb Image Browser, large image databases are visualised on a hexagonal lattice with image thumbnails occupying hexagons. Arranged in a space filling manner, visually similar images are located close together enabling large image datasets to be navigated in a hierarchical manner. Various browsing tools are incorporated to allow for interactive exploration of the database. Experimental results confirm that our approach affords efficient image retrieval. © 2010 IEEE.

Brazilian flavonoid database: Application of quality evaluation system

Much information on flavonoid content of Brazilian foods has already been obtained; however, this information is spread in scientific publications and non-published data. The objectives of this work were to compile and evaluate the quality of national flavonoid data according to the United States Department of Agriculture`s Data Quality Evaluation System (USDA-DQES) with few modifications, for future dissemination in the TBCA-USP (Brazilian Food Composition Database). For the compilation, the most abundant compounds in the flavonoid subclasses were considered (flavonols, flavones, isoflavones, flavanones, flavan-3-ols, and anthocyanidins) and the analysis of the compounds by HPLC was adopted as criteria for data inclusion. The evaluation system considers five categories, and the maximum score assigned to each category is 20. For each data, a confidence code (CC) was attributed (A, B, C and D), indicating the quality and reliability of the information. Flavonoid data (773) present in 197 Brazilian foods were evaluated. The CC ""C"" (as average) was attributed to 99% of the data and ""B"" (above average) to 1%. The main categories assigned low average scores were: number of samples; sampling plan and analytical quality control (average scores 2, 5 and 4, respectively). The analytical method category received an average score of 9. The category assigned the highest score was the sample handling (20 average). These results show that researchers need to be conscious about the importance of the number and plan of evaluated samples and the complete description and documentation of all the processes of methodology execution and analytical quality control. (C) 2010 Elsevier Inc. All rights reserved.