Articles in PDF available for free downloading (click on the headers):
This paper aims to introduce the problems and challenges concerned with the design and creation of CBIR systems, which is based on a free hand sketch (Sketched based image retrieval-SBIR). This analysis led us to studying the usability of a method for computing dissimilarity between user-produced pictorial queries and database images according to features extracted from Gray-Level Co-occurrence Matrix (GLCM) automatically.
CBIR is generally characterized by the methods that consumes less time. Hence fast content – based image retrieval is a need of the day especially image mining for shapes, as image database is growing exponentially in size with time. In this paper, texture features extracted from GLCM, tested, and investigated on different standard databases is proposed, it exhibits invariant to rotation. The retrieval performance of the proposed method is showed for both the dinosaurs retrieval efficiency achieved about 95% and precision also 95% where color is not dominant. It is also observed that the proposed method achieved low retrieval performance over these four image features for sketch based and color dominant images. This process can be used as coarse level in hierarchical CBIR that reduces the database size from very large set to a small one. This tiny database can further be scrutinized rigorously using the Edge Histogram Descriptor (EHD) and Color and Color Co-occurrence Matrix (CCM) etc.
A Content Based Image Retrieval (CBIR) system provides an efficient way of retrieving most similar images from image collections. In this paper we present a novel approach which combines color and edge features to extract similar images. We apply median filtering technique to original image to get the smoothed image. The Bi-directional Empirical Mode Decomposition (BEMD) technique is applied to extract edge information from the image. Then we replace only the values of edge position of smoothed image with the detected edge image values by BEMD and extracted 64 bins gray features. Later we apply one dimensional color histogram technique to obtain histogram vector by using RGB color space and is converted into 32 bins color features. Finally, we combine both the features to extract the most similar images from the database. The experiment is conducted on 1000 images of different categories stored in groundtruth database and the effectiveness of this technique is demonstrated. The results have been tabulated and compared with the conventional median and edge technique. We can observe that performance our proposed method is good.
LIRe (Lucene Image Retrieval) is an open source library for content based image retrieval. Besides providing multiple common and state of the art retrieval mechanisms it allows for easy use on multiple platforms. LIRe is actively used for research, teaching and commercial applications. Due to its modular nature it can be used on process level (e.g. index images and search) as well as on image feature level. Developers and researchers can easily extend and modify LIRe to adapt it to their needs.
Content-based image retrieval approach for biometric security using colour, texture and shape features controlled by fuzzy heuristics
In this paper, we discuss a new content-based image retrieval approach for biometric security, which is based on colour, texture and shape features and controlled by fuzzy heuristics. The proposed approach is based on the three well-known algorithms: colour histogram, texture and moment invariants. The use of these three algorithms ensures that the proposed image retrieval approach produces results which are highly relevant to the content of an image query, by taking into account the three distinct features of the image and similarity metrics based on Euclidean measure. Colour histogram is used to extract the colour features of an image. Gabor filter is used to extract the texture features and the moment invariant is used to extract the shape features of an image. The evaluation of the proposed approach is carried out using the standard precision and recall measures, and the results are compared with the well-known existing approaches. We present results which show that our proposed approach performs better than these approaches.
In this paper, we present a new and effective color image retrieval scheme for combining all the three i.e. color, texture and shape information, which achieved higher retrieval efficiency. Firstly, the image is predetermined by using fast color quantization algorithm with clusters merging, and then a small number of dominant colors and their percentages can be obtained. Secondly, the spatial texture features are extracted using a steerable filter decomposition, which offers an efficient and flexible approximation of early processing in the human visual system. Thirdly, the pseudo-Zernike moments of an image are used for shape descriptor, which have better features representation capabilities and are more robust to noise than other moment representations. Finally, the combination of the color, texture and shape features provide a robust feature set for image retrieval. Experimental results show that the proposed color image retrieval is more accurate and efficient in retrieving the user-interested images.
Research on image retrieval technology based on color feature, for the color histogram with a rotation, translation invariance of the advantages and disadvantages of lack of space, a color histogram and color moment combination Image Retrieval. The theory is a separate color images and color histogram moment of extraction, and then two methods of extracting color feature vector weighted to achieve similar distance, similar to the last distance based on the size of the return search results, based on the realization of the characteristics of the color image Retrieval system. The results show that the method is rotation, translation invariance, a single method of extracting color features, enhanced image search and improve the accuracy of the sort.
img(Anaktisi) is a C#/.NET content base image retrieval application suitable for the web. It provides efﬁcient retrieval services for various image databases using as a query a sample image, an image sketched by the user and keywords. The image retrieval engine is powered by innovative compact and effective descriptors. Also, an Auto Relevance Feedback (ARF) technique is provided to the user. This technique readjusts the initial retrieval results based on user preferences improving the retrieval score signiﬁcantly.
This paper presents an image retrieval suite called img(Rummager) which brings into effect a number of new as well as state of the art descriptors. The application can execute an image search based on a query image, either from XML-based index ﬁles, or directly from a folder containing image ﬁles, extracting the comparison features in real time. In addition the img(Rummager) application can execute a hybrid search of images from the application server, combining keyword information and visual similarity. Also img(Rummager) supports easy retrieval evaluation based on the normalized modiﬁed retrieval rank (NMRR) and average precision (AP).
The distribution of pixel colors in an image generally contains interesting information. Recently, many researchers have analyzed the color attributes of an image and used it as the features of the images for querying. Color histogram is one of the most frequently used image features in the field of color-based image retrieval. The color histogram is widely used as an important color feature indicating the contents of the images in content-based image retrieval (CBIR) systems. Specifically histogram-based algorithms are considered to be effective for color image indexing. Color histogram describes the global distribution of pixels of an image which is insensitive to variations in scale and easy to calculate. However, the high-resolution color histograms are usually high dimension and contain much redundant information which does not relate to the image contents, while the low-resolution histograms can not provide adequate discriminative information for image classification. And an image often includes a part of colors but not all, So there will be many accounts of colors are zeros. In order to save space, we shouldn’t need store them. In this paper, a color high-resolution, non-uniform quantized color histogram is proposed and the improving representation about histogram is proposed too. Major color, major segmentation block, and a new grayscale co-existing matrix’s method are proposed.
A study of order-based block color feature image retrieval compared with cumulative color histogram method
Color is one of the most important features in image retrieval. Color histograms have proved to be stable representations of an image, but they might be similar in different kinds of images because they describe the global intensity distribution of images. A new color image representation method is proposed in this paper. At first, each color channel (R, G, B) of an image is divided into 48 blocks (6 rows × 8 columns). Secondly, statistical features are computed to characterize the block’s color feature. Finally, all block features are combined to form an image’s color feature. The experimental results show that the retrieval effectiveness of the proposed technique is better than color histogram-based method.
The attempt to combine image structure and color information together for improving objective image quality assessment performance is introduced in this paper. Through analyzing the respective limitation of both SSIM, which focus on image structure, and CD-PSNR, which is based on S-CIELAB and works with image color only, the additive model (SAC) and the multiplicative model (SMC) are defined respectively. Our experimental results show that the proposed combined models all perform better than single method, and we could have different choices considering different purpose and the tradeoff between complexity and efficiency.
<em>Purpose</em> – The main obstacle in realising semantic-based image retrieval from the web is that it is difficult to capture semantic description of an image in low-level features. Text-based keywords can be generated from web documents to capture semantic information for narrowing down the search space. The combination of keywords and various low-level features effectively increases the retrieval precision. The purpose of this paper is to propose a dynamic approach for integrating keywords and low-level features to take advantage of their complementary strengths.
<em>Design/methodology/approach</em> – Image semantics are described using both low-level features and keywords. The keywords are constructed from the text located in the vicinity of images embedded in HTML documents. Various low-level features such as colour histograms, texture and composite colour-texture features are extracted for supplementing keywords.
<em>Findings</em> – The retrieval performance is better than that of various recently proposed techniques. The experimental results show that the integrated approach has better retrieval performance than both the text-based and the content-based techniques.
<em>Research limitations/implications</em> – The features of images used for capturing the semantics may not always describe the content.
Practical implications – The indexing mechanism for dynamically growing features is challenging while practically implementing the system.
<em>Originality/value</em> – A survey of image retrieval systems for searching images available on the internet found that no internet search engine can handle both low-level features and keywords as queries for retrieving images from WWW so this is the first of its kind.
In this paper, three image features are proposed for image retrieval. In addition, a feature selection technique is also brought forward to select optimal features to not only maximize the detection rate but also simplify the computation of image retrieval. The first and second image features are based on color and texture features, respectively called color co-occurrence matrix (CCM) and difference between pixels of scan pattern (DBPSP) in this paper. The third image feature is based on color distribution, called color histogram for K-mean (CHKM).
CCM is the conventional pattern co-occurrence matrix that calculates the probability of the occurrence of same pixel color between each pixel and its adjacent ones in each image, and this probability is considered as the attribute of the image. According to the sequence of motifs of scan patterns, DBPSP calculates the difference between pixels and converts it into the probability of occurrence on the entire image. Each pixel color in an image is then replaced by one color in the common color palette that is most similar to color so as to classify all pixels in image into k-cluster, called the CHKM feature.
Difference in image properties and contents indicates that different features are contained. Some images have stronger color and texture features, while others are more sensitive to color and spatial features. Thus, this study integrates CCM, DBPSP, and CHKM to facilitate image retrieval. To enhance image detection rate and simplify computation of image retrieval, sequential forward selection is adopted for feature selection. Besides, based on the image retrieval system (CTCHIRS), a series of analyses and comparisons are performed in our experiment. Three image databases with different properties are used to carry out feature selection. Optimal features are selected from original features to enhance the detection rate.
To begin with texture feature extraction, it present show to efficiently extract texture features to describe the content of images, then based on the common methods to acquire the fractal dimension of an image nowadays to extracted texture features, a new efficient algorithms of texture feature of two different fractal dimension is designed and realized in image retrieval system. The experimental results proved that fractal dimension as texture feature algorithm reflects texture feature of image more precisely. Conclusion: Be a large of quantity image retrieval, effective image feature extraction is a powerful tool in the research information classification and identification of areas with potential.
Developing methods for medical image characterization and indexing are in great demand for organizing and retrieving images from huge medical image databases. In this paper, novel algorithm based on tree structured cosine modulated wavelet transform (TSCMWT) for retrieval of Computer Tomography (CT) images is proposed. The proposed method performs better than existing available methods. Experimental results are promising in order to meet the requirements of the fast-paced clinical environment.
Based on color histogram, a new search method is presented in this paper. Firstly, quantify the image in HSV space and split it into annular isometric region. Calculate the third-order moment as feature vectors and realize image retrieval by the similarity comparing between the images. Experimental results show that this improved approach is better than the simple use of color histogram in an accurate survey.
A novel image feature called color variances among adjacent objects (CVAAO) is proposed in this study. Characterizing the color variances between contiguous objects in an image, CVAAO can effectively describe the principal colors and texture distribution of the image and is insensitive to distortion and scale variations of images. Based on CVAAO, a CVAAO-based image retrieval method is constructed. When given a full image, the CVAAO-based image retrieval method delivers the database images most similar to the full image to the user. This paper also presents a CVAAO-based ROI image retrieval method. When given a clip, the CVAAO-based ROI image retrieval method submits to the user a database image containing a target region most similar to the clip. The experimental results show that the CVAAO-based ROI image retrieval method can offer impressive results in finding out the database images that meet user requirements.
Due to its simple, low computation complexity and invariant to geometric transform, color feature is widely used in Content-based image retrieval (CBIR). In this paper, a new fuzzy method of color histogram is proposed based on the Lab color space, which links the three histograms created by the three components of the color space and provides a histogram which contains only ten bins. The experimental results prove that the proposed method is more accurate and robust than other methods of histogram.
This paper presents a new similarity measure method on a combination of color and texture feature representations. In this method, the YIQ color space is chosen, because it can describe both color images and gray images and the transform from RGB to YIQ is linear and simple than other color space. In the proposed method, we firstly segment image using texture feature by combination of wavelet transform and texture co-occurrence matrix and then quantize color feature in YIQ color space for every segme- ntation partition. Based on image segmentation and color quantization, a new kind of similarity measure is proposed. Compared with the traditional image retrieval methods, the proposed method is very efficient for the image retrieval purpose.
This project presents an approach to index and retrieve images using a compact color descriptor in two color spaces namely HSV,and CIE L*u*v*. A compact color descriptor adopted in the proposed content-based image retrieval system is 127-bit binary Haar color histogram, which is used as an index of the images in the database. The color histogram is obtained for an image using a suitable color space. Here, HSV color space, and CIE L*u*v*color space are chosen for comparison. A compact color descriptor is obtained from Haar transform coefficient of the color histogram. Each coefficient is quantized to a binary number, resulting in a 127-bit color descriptor. The descriptors of the images in the database are compared to the descriptor of the query image. The most similar images are retrieved and ordered according to their distances to the query. The proposed retrieval system can effectively retrieve the most similar images in CIE L*u*v* color space.
Color, texture and shape information have been the primitive image descriptors in content based image retrieval systems. This paper presents a novel framework for combining all the three i.e. color, texture and shape information, and achieve higher retrieval efficiency. The image is partitioned into non- overlapping tiles of equal size. The color moments and moments on gabor filter responses of these tiles serve as local descriptors of color and texture respectively. This local information is captured for two resolutions and two grid layouts that provide different details of the same image. An integrated matching scheme, based on most similar highest priority (MSHP) principle and the adjacency matrix of a bipartite graph formed using the tiles of query and target image, is provided for matching the images. Shape information is captured in terms of edge images computed using Gradient Vector Flow fields. Invariant moments are then used to record the shape features. The combination of the color, texture and shape features provide a robust feature set for image retrieval. The experimental results demonstrate the efficacy of the method.
The main idea of content-based image retrieval (CBIR) is to search on an image’s visual content directly. Typically, features (e.g., color, shape, texture) are extracted from each image and organized into a feature vector. Retrieval is performed by image example where a query image is given as input by the user and an appropriate metric is used to find the best matches in the corresponding feature space. We attempt to bypass the feature selection step (and the metric in the corresponding feature space) by following what we believe is the logical continuation of the CBIR idea of searching visual content directly. It is based on the observation that, since ultimately, the entire visual content of an image is encoded into its raw data (i.e., the raw pixel values), in theory, it should be possible to determine image similarity based on the raw data alone. The main advantage of this approach is its simplicity in that explicit selection, extraction, and weighting of features is not needed. This work is an investigation into an image dissimilarity measure following from the theoretical foundation of the recently proposed normalized information distance (NID) [M. Li, X. Chen, X. Li, B. Ma, P. Vitányi, The similarity metric, in: Proceedings of the 14th ACM-SIAM Symposium on Discrete Algorithms, 2003, pp. 863–872]. Approximations of the Kolmogorov complexity of an image are created by using different compression methods. Using those approximations, the NID between images is calculated and used as a metric for CBIR. The compression-based approximations to Kolmogorov complexity are shown to be valid by proving that they create statistically significant dissimilarity measures by testing them against a null hypothesis of random retrieval. Furthermore, when compared against several feature-based methods, the NID approach performed surprisingly well.
In this paper, we present a method using phase features for content based image retrieval (CBIR). Two related key issues of CBIR are feature extraction and similarity measure. However, most traditional methods treat them respectively and prevent further performance improvement. The method proposed here is based on the multi-scale local phase feature (MLPF) and local weighted phase correlation which combines the above two issues together by phase. And phase data is often locally stable with respect to noise, scale change and common illumination change. Moreover, we implement steerable filters to obtain rotation invariant. Finally, experiments have been conducted on image retrieval to show the effectiveness of the proposed method.
In order to improve the retrieval performance of images, this paper proposes an efficient approach for extracting and retrieving color images. The block diagram of our proposed approach to content-based image retrieval (CBIR) is given firstly, and then we introduce three image feature extracting arithmetic including color histogram, edge histogram and edge direction histogram, the histogram Euclidean distance, cosine distance and histogram intersection are used to measure the image level similarity. On the basis of using color and texture features separately, a new method for image retrieval using combined features is proposed. With the test for an image database including 766 general-purpose images and comparison and analysis of performance evaluation for features and similarity measures, our proposed retrieval approach demonstrates a promising performance. Experiment shows that combined features are superior to every single one of the three features in retrieval.
The field of color image retrieval has been an important research area for several decades. For the purpose of effectively retrieving more similar images from the digital image databases, this paper uses the color distributions, the mean value and the standard deviation, to represent the global characteristics of the image. Moreover, the image bitmap is used to represent the local characteristics of the image for increasing the accuracy of the retrieval system. As the experimental results indicated, the proposed technique indeed outperforms other schemes in terms of retrieval accuracy and category retrieval ability. Furthermore, the total memory space for saving the image features of the proposed method is less than Chan and Liu’s method.
A novel approach for texture image retrieval is proposed by using a new set of two-dimensional (2-D) rotated wavelet filters (RWF) and discrete wavelet transform (DWT) jointly. A new set of 2-D rotated wavelet improves characterization of diagonally oriented textures. Experimental results indicate that the proposed method improves retrieval rate from 70.09% to 78.44% on database D1, and from 75.62% to 80.78% on database D2, compared with the traditional DWT based approach. The proposed method also retains comparable levels of computational complexity.
Color histogram is now widely used in image retrieval. Color histogram-based image retrieval methods are simple and efficient but without considering the spatial distribution information of the color. To overcome the shortcoming of conventional color histogram-based image retrieval methods, an image retrieval method based on Radon Transform (RT) is proposed. In order to reduce the computational complexity, wavelet decomposition is used to compress image data. Firstly, images are decomposed by Mallat algorithm. The low-frequency components are then projected by RT to generate the spatial color feature. Finally the moment feature matrices which are saved along with original images are obtained. Experimental results show that the RT based retrieval is more accurate and efficient than traditional color histogram-based method in case that there are obvious objects in images. Further more, RT based retrieval runs significantly faster than the traditional color histogram methods.
We tackle the problem of retrieving images from a database. In particular we are concerned with the problem of retrieving images of airplanes belonging to one of the following six categories: 1) commercial planes on land, 2) commercial planes in the air, 3) war planes on land, 4) war planes in the air, 5) small aircrafts on land, and 6) small aircrafts in the air. During training, a wavelet-based description of each image is first obtained using Daubechies 4-wavelet transformation. The resulting coefficients are then used to train a neural network. During classification, test images are presented to the trained system. The coefficients are obtained from the Daubechies transform from histograms of a decomposition of the image into square sub-images of each channel of the original image. 120 images were used for training and 240 for independent testing. An 88% correct identification rate was obtained.
The novel approach combines colour and texture features for content based image retrieval. Features like colour and texture are obtained by computing the measure of standard deviation in combination with energy on each colour band of image and sub band of wavelet. Wavelet transform is used for decomposing the image into 2x2 subbands. Feature database in content-based image retrieval of 640 Visual texture (VisTex) color images is constructed. It is observed that proposed method outperforms the other conventional histograms and standard wavelet decomposition techniques.
This paper presents an efficient spatial indexing technique based on Silhouette moments that makes the index robust subject to the three basic transformations for CBIR. Spatial index is generated based upon a fast and robust clustering technique, which can recognize color clusters of any shape. The new clustering technique has been found to be efficient in terms of time complexity and cluster quality than many of its counterparts. A matching engine has been devised to retrieve images from the image database, which has the capacity for global and regional similarity search.
Color histogram is an important technique for color image database indexing and retrieving. However, the main problem with color histogram indexing is that it does not take the color spatial distribution into consideration. Previous researches have proved that the effectiveness of image retrieval increases when spatial feature of colors is included in image retrieval. In this paper, two new descriptors, color distribution entropy (CDE) and improved CDE (I-CDE), which introduce entropy to describe the spatial information of colors, are presented. In comparison with the spatial chromatic histogram (SCH) and geostat which also measure the global spatial relationship of colors, the experiment results show that CDE and I-CDE give better performance than SCH and geostat.
Image retrieval has emerged as an important problem in multimedia database management. This paper uses the color distribution, the mean value and the standard deviation, of an image as global information for image retrieval. Furthermore, this paper uses the common bitmap to represent the local characteristics of the image. The performance of the method is tested on three different image databases consisting of 410, 235, and 10,235 images. The third database has been partitioned into 10 categories for exploring the category retrieval ability. According to the experimental results, we find that the proposed method can effectively retrieve more similar images than other methods and the category ability is also higher than others. In addition, the total memory space for saving the image features of the proposed method is less than other methods.
This paper describes a new color image segmentation method based on low-level features including color, texture and spatial information. The mean-shift algorithm with color and spatial information in color image segmentation is in general successful, however, in some cases, the color and spatial information are not sufficient for superior segmentation. The proposed method addresses this problem and employs texture descriptors as an additional feature. The method uses wavelet frames that provide translation invariant texture analysis. The method integrates additional texture feature to the color and spatial space of standard mean-shift segmentation algorithm. The new algorithm with high dimensional extended feature space provides better results than standard mean-shift segmentation algorithm as shown in experimental results.
One of the features in JPEG2000 is ROI (region of interest) coding technique. Since the shape of interested region is manually optional in the coding process, the disturbance of uninterested regions to the retrieval process could be controlled to be very small if we retrieve images based on ROI content. This paper presents a novel and effective scheme for remote sensing image retrieval, which does not need to decode JPEG2000's code stream completely. We extract the spectral features of the objects based on the properties that objects would reflect different waves in different wave bands. The subsequent retrieval is based on this kind of spectral features. In addition, we design a new measurement scheme by which similarity between two images is computed and then the retrieval is realized based on the measurement. Experimental results show that our method is accurate and efficient. It also shows obviously that our method costs much less time than the traditional ones.
Searching for similar images is an important research topic for multimedia database management. This paper uses a quality index model to search for similar images from digital image databases. In order to speed up retrieval, the quality index model is partitioned into three factors: loss of correlation, luminance distortion, and contrast distortion. The method is performed on three different image databases to test for retrieval accuracy and category retrieval ability. The experimental results show that the proposed method performs better than the color histogram method, the color moment method, and the CDESSO method.
In this study, a conceptually simple, yet flexible and extendable strategy to contrast two different color images is introduced. The proposed approach is based on the multivariate Wald-Wolfowitz test, a nonparametric test that assesses the commonality between two different sets of multivariate observations. It provides an aggregate gauge of the match between color images, taking into consideration all the (selected) low-level characteristics, while alleviating correspondence issues. We show that a powerful measure of similarity between two color images can emerge from the statistical comparison of their representations in a properly formed feature space. For the sake of simplicity, the RGB-space is selected as the feature space, while we are experimenting with different ways to represent the images within this space. By altering the feature-extraction implementation, complementary ways to portray the image content appear. The reported results, from the application on a diverse collection of images, clearly demonstrate the effectiveness of our method, its superiority over previous methods, and suggest that even further improvements can be achieved along the same line of research. It is not only the unifying character that makes our strategy appealing, but also the fact that the retrieval performance does not increase continuously with the amount of details in the image representation. The latter sets an upper limit to the computational demands and reminds of performance plateaus reached by novel approaches in information retrieval.
This paper focuses on developing a Fast And Semantics-Tailored (FAST) image retrieval methodology. Specifically, the contributions of FAST methodology to the CBIR literature include: (1) development of a new indexing method based on fuzzy logic to incorporate color, texture, and shape information into a region-based approach to improving the retrieval effectiveness and robustness; (2) development of a new hierarchical indexing structure and the corresponding hierarchical, elimination-based A* retrieval (HEAR) algorithm to significantly improve the retrieval efficiency without sacrificing the retrieval effectiveness; it is shown that HEAR is guaranteed to deliver a logarithm search in the average case; (3) employment of user relevance feedback to tailor the effective retrieval to each user's individualized query preference through the novel indexing tree pruning (ITP) and adaptive region weight updating (ARWU) algorithms. Theoretical analysis and experimental evaluations show that FAST methodology holds great promise in delivering fast and semantics-tailored image retrieval in CBIR.
Content-based image retrieval (CBIR) is a collection of techniques for retrieving images on the basis of features, such as color, texture and shape. An efficient tool, which is widely used in CBIR, is that of color image histograms. The classic method of color histogram creation results in very large histograms with large variations between neighboring bins. Thus, small changes in the image might result in great changes in the histogram. Moreover, the fact that each color space consists of three components leads to 3-dimensional histograms. Manipulating and comparing 3D histograms is a complicated and computationally expensive procedure. The need, therefore, for reduction of the three dimensions to one could lead to efficient approaches. This procedure of projecting the 3D histogram onto one single-dimension histogram is called histogram linking. In this paper, a new fuzzy linking method of color histogram creation is proposed based on the L*a*b* color space and provides a histogram which contains only 10 bins. The histogram creation method in hand was assessed based on the performances achieved in retrieving similar images from a widely diverse image collection. The experimental results prove that the proposed method is less sensitive to various changes in the images (such as lighting variations, occlusions and noise) than other methods of histogram creation.
This paper describes a content-based approach to improve image retrieval effectiveness. First, we define two new measures for computing similarity among images based on color histograms, namely the dissimilitude distance <i>DS</i>* and the similarity distance <i>E</i>. The latter is incorporated into the exponentiation part of the Gibbs distribution and into the generalized Dirichlet mixture, while the former is compared to five similarity measures: <i>L</i><sub>1</sub>, <i>L</i><sub>2</sub> (Euclidean distance), <i>E</i> as well as Gibbs and Dirichlet distributions integrating the similarity measure <i>E</i>. Then, in order to overcome the limitations (and inappropriateness) of some previous information retrieval measures in evaluating the efficiency of an image retrieval process, three variants of a new effectiveness measure are proposed and experimented on an image collection for different similarity distances.
Image retrieval using histograms of uni-color and bi-color blocks and directional changes in intensity gradient
In this paper, we present a new and effective image indexing technique that employs local uni-color and bi-color distributions and local directional distribution of intensity gradient. The image is divided into 4 by 4 non-overlapping blocks. Each block, based on its gradient magnitude, is classified as uniform or non-uniform. Using the average of each color component for the pixels of a uniform block, its representative color is found. Then the histogram of uni-color uniform blocks of the image, HUCUB, is constructed. To each non-uniform block, two representative colors are assigned. Then the histogram of bi-color non-uniform blocks, HBCNB, is created. To represent the shape content of the image, the histogram of directional changes in intensity gradient, HDCIG, is introduced. Experimental results on a database of 2250 images are reported.
Fast indexing method for image retrieval using k nearest neighbors searches by principal axis analysis
This paper presents a fast indexing scheme for content-based image retrieval based on the principal axis analysis. Image databases often represent the image objects as high-dimensional feature vectors and access them via the feature vectors and similarity measure. A similarity measure similar to the quadratic histogram distance measure is defined for this indexing method. The computational complexity of similarity measure in high-dimensional image database is very huge and hence the applications of image retrieval are restricted to certain areas. In this work, feature vectors in a given image are ordered by the principal axis analysis to speed up the similarity search in a high-dimensional image database using k nearest neighbor searches. To demonstrate the effectiveness of the proposed algorithm, we conducted extensive experiments and compared the performance with the IBM’s query by image content (QBIC) method, Jain and Vailay’s method, and the LPC-file method. The experimental results demonstrate that the proposed method outperforms the compared methods in retrieval accuracy and execution speed. The execution speed of the proposed method is much faster than that of QBIC method and it can achieve good results in terms of retrieval accuracy compared with Jain’s method and QBIC method.
Feature extraction is one of the most important tasks for efficient and accurate image retrieval purpose. In this paper we have presented a Cosine-modulated wavelet transform based technique for extraction of texture features. The major advantages of Cosine-modulated wavelet transform are less implementation complexity, good filter quality, and ease in imposing the regularity conditions. Texture features are obtained by computing the energy, standard deviation and their combination on each subband of the decomposed image. To check the retrieval performance, texture database of 1856 textures is created from Brodatz album. Retrieval efficiency and accuracy using Cosine-modulated wavelet based features is found to be superior to other existing methods.
We present a new technique for content based image retrieval using motif cooccurrence matrix (MCM). The MCM is derived from the motif transformed image. The whole image is divided into 2×2 pixel grids. Each grid is replaced by a scan motif that minimizes the local gradient while traversing the 2×2 grid forming a motif transformed image. The MCM is then defined as a 3D matrix whose (i,j,k) entry denotes the probability of finding a motif i at a distance k from the motif j in the transformed image. Conceptually, the MCM is quite similar to the color cooccurrence matrix (CCM), however, the retrieval using the MCM is better than the CCM since it captures the third order image statistics in the local neighborhood. Experiments confirm that the use of MCM considerably improves the retrieval performance.
A content based image retrieval (CBIR) system retrieves relevant images from an image database. Over the years, several methods have been proposed to extract these features. Previous researches show that the effectiveness of a CBIR system increases when spatial relationship of colours is considered. In this paper we propose using the Looseness parameter from geostat, a branch of statistics which deals with geographical data, to describe the global spatial relationship of colours. Spatial chromatic histogram (SCH) is another method which also measures the global spatial relationship of colours. However, the spatial measurement of SCH is size variant, the spatial measurement of geostat is size invariant. We analyse and compare the performance of geostat and SCH.
Colour is one of the most important features in content based image retrieval. However, colour is rarely used as a feature that codes local spatial information, except for colour texture. This paper presents an approach to represent spatial colour distributions using local principal component analysis (PCA). The representation is based on image windows which are selected by two complementary data driven attentive mechanisms: a symmetry based saliency map and an edge and corner detector. The eigenvectors obtained from local PCA of the selected windows form colour patterns that capture both low and high spatial frequencies, so they are well suited for shape as well as texture representation. Projections of the windows selected from the image database to the local PCs serve as a compact representation for the search database. Queries are formulated by specifying windows within query images. System feedback makes both the search process and the results comprehensible for the user.
A novel algorithm based on running sub-blocks with different similarity weights is proposed for object-based image retrieval. By splitting the entire image into certain sub-blocks, we use color region information and similarity matrix analysis to retrieval images under the query of special object.
In many colour-based image retrieval systems the colour properties of an image are described by its colour histogram. Histogram-based search is, however, often inefficient for large histogram sizes. Therefore we introduce several new, Karhunen–Loève transform (KLT)-based methods that provide efficient representations of colour histograms and differences between two colour histograms. The methods are based on the following two observations; Ordinary KLT considers colour histograms as signals and uses the Euclidian distance for optimization; KLT with generalized colour distance measures that take into account both the statistical properties of the image database and the properties of the underlying colour space should improve the retrieval performance. Image retrieval applications compare similarities between different images. Relevant for the decision is only the local structure of the image space around the current query image since the task is to find those images in the database that are most similar to this given query image. Therefore only the local topology of the feature space is of interest and compression methods should preserve this local topology as much as possible. It is therefore more important to have a good representation of the differences between features of similar images than good representations of the features of the images themselves. The optimization should therefore be based on minimizing the approximation error in the space of local histogram differences instead of the space of colour histograms. In this paper we report the results of our experiments that are done on three image databases containing more than 130,000 images. Both objective and subjective ground truth queries are used in order to evaluate the proposed methods and to compare them with other existing methods. The results from our experiments show that compression methods based on a combination of the two observations described above provide new, powerful and efficient retrieval algorithms for colour-based image retrieval.
Content-based image retrieval systems have become a reliable tool for many image database applications. There are several advantages of the image retrieval techniques compared to other simple retrieval approaches such as text-based retrieval techniques. This paper proposes a new image retrieval technique that can be used for retrieving color images. The proposed technique is based on a fractal scanning procedure, which extracts 1-D signatures for each one of the image color components. These signatures contain not only color information, but also shape and textural image information. Using Fourier descriptors and discrete transform, powerful features are extracted from the signatures that permit the efficient retrieval of color images. The system is suitable for retrieving query images even in distortion cases such as deformations, noise, color, cosine reduction and smoothing.
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.
This chapter summarises the current state of the art in content based image retrieval (CBIR). It discusses the need for image retrieval by content, and the types of query which might be encountered. It describes the main techniques currently used to retrieve images by content at both primitive and semantic levels, describes the features of some commercial and experimental CBIR systems, assesses the capabilities of current technology, and outlines possible future developments the field.
The paper presents a review of 200 references in content-based image retrieval. The paper starts with discussing the working conditions of content-based retrieval: patterns of use, types of pictures, the role of semantics, and the sensory gap. Subsequent sections discuss computational steps for image retrieval systems. Step one of the review is image processing for retrieval sorted by color, texture, and local geometry. Features for retrieval are discussed next, sorted by: accumulative and global features, salient points, object and shape features, signs, and structural combinations thereof. Similarity of pictures and objects in pictures is reviewed for each of the feature types, in close connection to the types and means of feedback the user of the systems is capable of giving by interaction. We briefly discuss aspects of system engineering: databases, system architecture, and evaluation. In the concluding section, we present our view on: the driving force of the field, the heritage from computer vision, the influence on computer vision, the role of similarity and of interaction, the need for databases, the problem of evaluation, and the role of the semantic gap.
This paper presents a Content Based System Retrieval that uses gradient color fields as features. These features take into account both contour curvature and colors found in adjacent regions. This approach is dual from color region based methods. It allows a simple description of images using a coarse understandable sketch, which preserves rich information with a small amount of storage. This representation allows different type of queries such as: statistical or structural, global or partial and contour based. These features have been tested, using different modes on a very large image database from broadcast television. Results obtained from our system is presented and discussed.
The problem of computing the similarity between two images is transformed to that of approximating the distance between two extended region adjacency graphs, which are extracted from the images in time and space linear in the number of pixels. Invariance to translation and rotation is thus achieved. Invariance to scaling is also achieved by taking the relative size of regions into account. Furthermore, the method provides a trade-off between pixel similarity threshold and approximation of the distance measure, which can be used to bound the error in image recognition as well as the time complexity of the computation.
The paper describes a new indexing methodology for image databases integrating color and spatial information for content-based image retrieval. This methodology, called Spatial-Chromatic Histogram (SCH), synthesizing in few values information about the location of pixels having the same color and their arrangement within the image, can be more satisfactory than standard techniques when the user would like to retrieve from the database the images that actually resemble the query image selected in their color distribution characteristics. Experimental trials on a database of about 3000 images are reported and compared with more standard techniques, like Color Coherence Vectors, on the basis of human perceptual judgments.
One of the challenging problems for image content retrieval is how to represent semantics of images. We have proposed a new representation for image contents using visual ontologies and wavelets. We will discuss how to represent the shape of objects/regions of interest, colour distribution and texture of the images integrating visual ontologies with wavelets. Our ideas are: (1) propose a set of meta ontologies and semantic descriptors (graphic and symbolic) for visual ontologies; (2) using visual ontologies and most significant coefficients of a 2D multiresolution wavelet transform to represent image contents. The visual ontologies are represented by a set of descriptors, relations, logic operators, and functions; (3) using a multiresolution wavelet transform to extract the colour and texture features of target images and map it into relevant visual ontologies. The query images can also be represented as visual ontologies or most significant coefficients of a 2-D multiresolution wavelet; (4) a spatial query can be integrated with other visual ontologies. The initial experiments have shown encouraging results.
Most image-retrieval systems rely on similarity measures for collecting images of similar types. Similarity measures are an integral part in the development of image management systems. In this paper, we propose frame-based similarity measures for accessing structured images, e.g. images can be understood by inferring from objects present and the relationships among them. The image content is described in the following ways: (1) adjacency blocks and/or (2) unary and binary attributes that are used to fill frame slots for representing image structures. The retrieval is based on similarity measures by comparing the contents of the query image and database image. The frame-based representation scheme is application-independent. Our similarity measures allow for images to be retrieved with different degrees of similarity and are flexible. We have developed a prototype system using the paradigms proposed. We demonstrate the usefulness of our system with some experimental results.
Digital image indexing and retrieval by content using the fractal transform for multimedia databases
Digital image database represent huge amount of data, automatic indexing and content base retrieval are crucial factors.Content base retrieval requests specific indexing techniques allowing to deal with efficient data representations (reduced amount of data and adapted to content base retrieval) in order to avoid prohibitive low level process time on queries. We present a method consisting in building an “iconic” level image representation preserving the semantic, allowing to process content base retrieval and to reconstruct the image. This representation consists in a set of function parameters.It is a semantic preserving compression built by using a dedicated Fractal compress scheme. To search patterns in an huge set of images, a specific algorithm allows to use this “iconic” representation as an index.It works entirely in the Fractal transform parameter space of both image and pattern, to obtain performances compatible with an interactive search. The research engine uses both textures and edges of the pattern. The pattern can be present in the image with different orientations and/or scales by using a multiresolution Fractal representation of the pattern. This method allows to retrieve in 3 seconds a 64×64 pixels pattern in an 100 images (512×512 pixels) database, on a SUN Sparc 20 workstation.It can be combined with other indexing and retrieval techniques, such as textual annotation.
This paper deals with efficient retrieval of images from large databases based on the color and shape content in images. With the increasing popularity of the use of large-volume image databases in various applications, it becomes imperative to build an automatic and efficient retrieval system to browse through the entire database. Techniques using textual attributes for annotations are limited in their applications. Our approach relies on image features that exploit visual cues such as color and shape. Unlike previous approaches which concentrate on extracting a single concise feature, our technique combines features that represent both the color and shape in images. Experimental results on a database of 400 trademark images show that an integrated color- and shape-based feature representation results in 99% of the images being retrieved within the top two positions. Additional results demonstrate that a combination of clustering and a branch and bound-based matching scheme aids in improving the speed of the retrievals.
Color is an important attribute for image matching and retrieval. We present a new method fo color matching based on a clustering algorithm in 3-D color space. We define a new color feature to characterize the color information and a distance measure to compute the color similarity of images. We have implemented this technique and tested it for a database of approximately 170 images. The test results shoe that the “Efficiency of Retrieval” of this new method is very high.