Image indexing and retrieval based on color histograms

The image data has grown fast in fields like social media, healthcare, and e-commerce and this has enhanced the demand of quick and precise image retrieval systems. Conventional methods of retrieval that have relied on handcrafted methods of feature extraction and pixel-wise comparison are costly in computer time and cannot withstand changes in image quality, lighting, and view point. To resolve these drawbacks, this project adopts an effective content-based image retrieval framework with a Vision Transformer (ViT) to extract features and FAISS (Facebook AI Similarity Search) to conducted fast similarity search. The system gets compact numerical representations of features of images as a result of a deep learning model and constructs a scale index to search similarity. The FAISS IndexIVFFlat architecture allows to search the nearest-neighbor approximationally with poor scale and low query time by grouping image features into clusters and lessening the search space at query time. Metadata of the indexed images are stored in the system and the top-k most similar images are retrieved according to the distance scores. This method is accurate and fast thus appropriate when dealing with large scale image databases. The system relies on the IndexIVFFlat structure of FAISS to cluster image features into regions which is a significant reduction of the search space that can support meaningful approximate nearest-neighbor queries. Results indicate that the system can be effective in separating useful and useless images that have high similarity (up to 0.819) and low distances (0.220) that ensure the accuracy and scalability of the system to image databases of large scales.

BackgroundThere is increasing interest in the development of computational methods to analyze fluorescent microscopy images and enable automated large-scale analysis of the subcellular localization of proteins. Determining the subcellular localization is an integral part of identifying a protein's function, and the application of bioinformatics to this problem provides a valuable tool for the annotation of proteomes. Training and validating algorithms used in image analysis research typically rely on large sets of image data, and would benefit from a large, well-annotated and highly-available database of images and associated metadata.DescriptionThe Yeast Resource Center Public Image Repository (YRC PIR) is a large database of images depicting the subcellular localization and colocalization of proteins. Designed especially for computational biologists who need large numbers of images, the YRC PIR contains 532,182 TIFF images from nearly 85,000 separate experiments and their associated experimental data. All images and associated data are searchable, and the results browsable, through an intuitive web interface. Search results, experiments, individual images or the entire dataset may be downloaded as standards-compliant OME-TIFF data.ConclusionsThe YRC PIR is a powerful resource for researchers to find, view, and download many images and associated metadata depicting the subcellular localization and colocalization of proteins, or classes of proteins, in a standards-compliant format. The YRC PIR is freely available at http://images.yeastrc.org/.

Image retrieval with relevant feedback on large and high-dimensional image databases is a challenging task. In this paper, we propose an image retrieval method, called BCFIR (Binary Codes for Fast Image Retrieval). BCFIR utilizes sparse discriminant analysis to select the most important original feature set, and solve the small class problem in the relevance feedback. Besides, to increase the retrieval performance on large-scale image databases, in addition to BCFIR mapping real-valued features to short binary codes, it also applies a bagging learning strategy to improve the ability general capabilities of autoencoders. In addition, our proposed method also takes advantage of both labeled and unlabeled samples to improve the retrieval precision. The experimental results on three databases demonstrate that the proposed method obtains competitive precision compared with other state-of-the-art image retrieval methods.

A Survey on Automatic Image Annotation and Trends of the New Age

This paper proposes a hybrid approach of ontology and image clustering to automatically generate hierarchic image database. In the field of computer vision, ”generic object recognition” is one of the most important topics. Generic object recognition needs three types of research: feature extraction, pattern recognition, and database preparation; this paper targets at database preparation. The proposed approach considers both object semantic and visual features in images. In the proposed approach, the semantic is covered by ontology framework, and the visual similarity is covered by image clustering based on Gaussian Mixture Model. The image database generated by the proposed approach covered over 4,800 concepts (where 152 concepts have more than 100 images) and its structure was hierarchic. Through the subjective evaluation experiment, whether images in the database were correctly mapped or not was examined. The results of the experiment showed over 84% precision in average. It was suggested that the generated image database was sufficiently practicable as learning database for generic object recognition.

Query by image content is a method to retrieve the most important images from the image database. It is an answer for the problem of searching for digital images in large database. A large number of relevance feedback schemes have been developed to improve the performance of content based image retrieval. In this paper we propose biased discriminant Euclidean embedding that form intraclass geometry and interclass discrimination. In this method images can be grouped by their similarity. In order to achieve this, firstly the visual features of the images are found out. In addition to this we use two clustering algorithms to assemble the images into clusters using their low level visual features. Here we have to filter the images in the hierarchical clustering and then apply the clustered images to K-Means, so that we can get the most relevant images. Query image is compared directly with the images in the clusters. Therefore number of comparisons is reduced because query image is compared only with the images in the clusters instead of comparing with all images in the database.

Color histogram is widely used for image indexing in content-based image retrieval (CBIR). A color histogram describes the global color distribution of an image. It is very easy to compute and is insensitive to small changes in viewing positions. However, the histogram is not robust to large appearance changes. Moreover, the histogram might give similar results for different kinds of images if the distributions of colors are same in the images. On the other hand, color correlogram is efficiently used for image indexing in content-based image retrieval. Color correlogram extracts not only the color distribution of pixels in images like color histogram, but also extracts the spatial information of pixels in the images. The characteristic of the color Correlogram to take into account the spatial information as well as the distribution of color pixels greatly attracts the researcher for content based image retrieval. In this paper, we propose the image bin (histogram value divisions) separation technique followed by extracting maxima of frequencies and plotting a correlogram. At first, the histogram is first calculated for an image. After that, it is subdivided into four equal bins. Each bin is subdivided into four more bins and for every such subdivision the maxima of frequencies s calculated. This information is stored in the form of a correlogram. The distance between correlogram of the query image with the corresponding correlogram of database images is calculated. The proposed algorithm is tested on a database comprising a large number of images.

A conventional color histogram (CCH) considers neither the color similarity across different bins nor the color dissimilarity in the same bin. Therefore, it is sensitive to noisy interference such as illumination changes and quantization errors. Furthermore, CCHs large dimension or histogram bins requires large computation on histogram comparison. To address these concerns, this paper presents a new color histogram representation, called fuzzy color histogram (FCH), by considering the color similarity of each pixel's color associated to all the histogram bins through fuzzy-set membership function. A novel and fast approach for computing the membership values based on fuzzy c-means algorithm is introduced. The proposed FCH is further exploited in the application of image indexing and retrieval. Experimental results clearly show that FCH yields better retrieval results than CCH. Such computing methodology is fairly desirable for image retrieval over large image databases.

Color histograms are widely used for content-based image retrieval because they are trivial and effective to compute, not sensitive to size and orientation, and robustly tolerate movement of objects in the image and changes in camera viewpoint. However, a color histogram only records an image's overall color compositions and no any spatial information is included, and also retrieval precision is limited. In this paper, a color-spatial histogram, which incorporates spatial information with color compositions without sacrificing the robustness of color histograms, is proposed. The purpose is considering the representation role of color compositions and spatial information in the meantime. Each entry in a color-spatial histogram is the color frequency and corresponding position information of the virtual boundary. Experimental evidence suggests that this new color-spatial histogram outperform not only the traditional color histogram method but also the other color-spatial histogram methods for image retrieval.

In many multi-media applications it is desirable to separate the influence of the illumination sources and imaging equipment from the properties of the depicted scene. The ability of the human visual system to solve this task in many situations is known as color constancy. Technical applications of these methods include automatic color correction and illumination independent search in image databases. Many conventional computational color constancy methods assume that the effect of an illumination change can be described by a matrix multiplication with a diagonal matrix. In this paper we introduce a color normalization algorithm which computes the unique color transformation matrix which normalizes a given set of moments computed from the color distribution of an image. This normalization procedure is a generalization of the channel independent color constancy methods since general matrix transformations are considered. We compare the performance of this new normalization method with conventional color constancy methods. The experiments show that diagonal transformation matrices provide a better illumination compensation. This shows that the color moments also contain significant information about the color distributions of the objects in the image which is independent of the illumination characteristics. In another set of experiments we use the unique transformation matrix as a descriptor of the set of moments which describe the global color distribution in the image. Combining the matrices computed from two such images describes the color differences between them. We then use this as a tool for color dependent search in image databases. This matrix based color search is computationally less demanding than histogram based color search tools.

Color histogram analysis is a powerful tool for characterizing color images. It has been widely used in image indexing and retrieval systems. A key problem to use color histogram in image classification is to find a robust similarity measurement between different color histograms. In this paper, we propose to use a cross-correlation function to measure color histogram similarity. We show that a cross-correlation function has several advantages over the method of histogram intersection, which has been widely used to calculate the similarity between color histograms: A cross-correlation function is normalized automatically; it can determine the similarity irrespective of image size; it is invariant to small color shift; it is easier to implement using the computationally efficient methods. We present an example of unsupervised image clustering by applying cross-correlation function to color histograms. This method was used to improve the perceived color consistency in a multi-print-engine system. We also show how to optimize the cross-correlation function to compensate for the color shift.

Contrast distortion is often a determining factor in human perception of image quality, but little investigation has been dedicated to quality assessment of contrast-distorted images without assuming the availability of a perfect-quality reference image. In this letter, we propose a simple but effective method for no-reference quality assessment of contrast distorted images based on the principle of natural scene statistics (NSS). A large scale image database is employed to build NSS models based on moment and entropy features. The quality of a contrast-distorted image is then evaluated based on its unnaturalness characterized by the degree of deviation from the NSS models. Support vector regression (SVR) is employed to predict human mean opinion score (MOS) from multiple NSS features as the input. Experiments based on three publicly available databases demonstrate the promising performance of the proposed method.

Content-based image retrieval is an active and fast advancing research area since the 1990s as a result of advances in the Internet and new digital image sensor technologies. However, many challenging research problems continue to attract researchers from multiple disciplines. Content-based image retrieval uses the visual contents of an image as features to represent and index the image to be searched from large scale image databases. The quality of the selected features relies mainly on the degree of the invariance property that is ensured under acceptable manipulations. This paper proposes an efficient method for compactly representing color and texture features and combining them for image retrieval. The performance of retrieval based on these compact descriptors obtained by the proposed techniques is analyzed and tested on wang database images yielding satisfactory accuracy rates. A comparative study demonstrated that the developed feature extraction scheme outperformed the other schemes being compared with.

In the past few years, Bag of Word (BOW) plays an important role in the field of image retrieval and automatic annotation especially in large scale image databases. However, in practice, the spatial location information of visual word is often ignored. Despite much previous work, it remains a challenging task to extract the spatial location information of visual word. Recently, salient object detection has been attracting lots of interest and obtains a series of satisfactory results. In this paper, the novel approach can encode spatial location information into standard BOW representation which utilize salient object in the image. The proposed approach is so simple, yet useful and efficient enough to be applied in large-scale databases. The main properties of this approach are: (1) the potential strength of visual word's spatial location information is extracted and a novel BOW scheme is proposed. (2) the advantage of standard BOW scheme is still retained such as rotational invariance, translational invariance and scaling invariance. (3) The new BOW scheme could significantly enhance the discriminative power. We evaluate our new framework on the standard dataset and the results have shown a significant improvement over the state-of-the-art methods.

Scene classification using semantic description has gained much attention towards automatic image retrieval. In many cases, visual appearance of images get affected by environmental conditions such as low lighting and viewing conditions. Such problems in semantic scenes pose difficult challenges during the classification of sceneries. To address this issue, a new outdoor scene classification method for using low level feature has been proposed in this work. To support automatic scene classification at the concept level an efficient illumination and rotation invariant low level features such as color, texture and edge like features have been used in conjunction with multiclass Support Vector Machine (SVM). In this work, we have taken scene categories like mountains, forests, highways, rivers, buildings etc., from the outdoor scenes for classification experimentation. From the experimental results, we demonstrate that the proposed method provides better classification in the large scale image databases like Eight scene category, upright scene and COREL dataset and gives better performance in terms of classification accuracy.