Papers by Roumiana Kountcheva
Method and system for digital watermarking of multimedia signals
Third-Order Tensor Representation Through Reduced Inverse Difference Pyramid
2019 International Conference on Creative Business for Smart and Sustainable Growth (CREBUS), 2019
In this work is presented a method for the third-order tensor representation through Reduced Inve... more In this work is presented a method for the third-order tensor representation through Reduced Inverse Difference Pyramid. In each of the pyramid levels is used the 3D Walsh-Hadamard transform and as a result is achieved high concentration of the tensor energy in a minimum number of spectrum coefficients, most of which - in the first decomposition level. The tensor is thus transformed into multi-layer spectrum tensor of same size. The corresponding decomposition pyramid is not "overcomplete", and is called „reduced". The representation has minimum computational complexity because the only operation needed for the execution, is „addition". The evaluation of the pyramid properties opens new abilities for its application in various areas, aimed at the information redundancy reduction in multidimensional signals and data.
Locally Adaptive Processing of Color Tensor Images Represented as Vector Fields
Smart innovation, systems and technologies, 2024
Deep Representation and Analysis of Visual Information, Based on the IDP Decomposition
Smart innovation, systems and technologies, 2024
Symmetry
A tensor representation structure based on the multilayer tensor spectrum pyramid (MLTSP) is intr... more A tensor representation structure based on the multilayer tensor spectrum pyramid (MLTSP) is introduced in this work. The structure is “truncated”, i.e., part of the high-frequency spectrum coefficients is cut-off, and on the retained low-frequency coefficients, obtained at the output of each pyramid layer, a hierarchical tensor SVD (HTSVD) is applied. This ensures a high concentration of the input tensor energy into a small number of decomposition components of the tensors obtained at the coder output. The implementation of this idea is based on a symmetrical coder/decoder. An example structure for a cubical tensor of size 8 × 8 × 8, which is represented as a two-layer tensor spectrum pyramid, where 3D frequency-ordered fast Walsh–Hadamard transform and HTSVD are used, is given in this paper. The analysis of the needed mathematical operations proved the low computational complexity of the new approach, due to a lack of iterative calculations. The high flexibility of the structure i...
In the paper is presented one new approach for creation and management of medical image databases... more In the paper is presented one new approach for creation and management of medical image databases with multi-layer access, based on the Inverse Pyramid Decomposition (IPD). The archived visual data is compressed using the special IDP format, presented in detail in this paper. The new approach offers flexible tools for multi-layer transfer of the processed information with consecutively quality improvement, together with reliable content protection ensured by digital watermark insertion and data hiding. The IDP permits insertion of multiple watermarks in same file. Part of the visual information (specific regions of interest in the medical images) is hidden for the low access levels and could be revealed by authorized users only.
Data Hiding with Image Decomposition Based on General Spectrum Pyramid
The paper presents a new method for image compression and hiding, based on decomposition with mul... more The paper presents a new method for image compression and hiding, based on decomposition with multi-layer general spectrum pyramid (GSP). The presented approach permits the insertion of one or more images between the consecutive decomposition layers. The additional information could be used for image authentication, for image contents protection or for providing additional confidential visual information to authorized users. In
New Approaches for Hierarchical Image Decomposition, Based on IDP, SVD, PCA and KPCA
Intelligent systems reference library, 2016
The contemporary forms of image representation vary depending on the application. There are well-... more The contemporary forms of image representation vary depending on the application. There are well-known mathematical methods for image representation, which comprise: matrices, vectors, determined orthogonal transforms, multi-resolution pyramids, Principal Component Analysis (PCA) and Independent Component Analysis (ICA), Singular Value Decomposition (SVD), wavelet sub-band decompositions, hierarchical tensor transformations, nonlinear decompositions through hierarchical neural networks, polynomial and multiscale hierarchical decompositions, multidimensional tree-like structures, multi-layer perceptual and cognitive models, statistical models, etc. In this chapter are analyzed the basic methods for hierarchical decomposition of grayscale and color images, and of sequences of correlated images of the kind: medical, multispectral, multi-view, etc. Here is also added one expansion and generalization of the ideas of the authors from their previous publications, regarding the possibilities for the development of new, efficient algorithms for hierarchical image decompositions with various purposes. In this chapter are presented and analyzed the following four new approaches for hierarchical image decomposition: the Branched Inverse Difference Pyramid (BIDP), based on the Inverse Difference Pyramid (IDP); the Hierarchical Singular Value Decomposition (HSVD) with tree-like computational structure; the Hierarchical Adaptive Principle Component Analysis (HAPCA) for groups of correlated images; and the Hierarchical Adaptive Kernel Principal Component Analysis (HAKPCA) for color images. In the chapter are given the algorithms, used for the implementation of these decompositions, and their computational complexity is evaluated. Some experimental results, related to selected applications are also given, and various possibilities for the creation of new hybrid algorithms for hierarchical decomposition of multidimensional images are specified. On the basis of the results obtained from the executed analysis, the basic application areas for efficient image processing are specified, such as: reduction of the information surplus; noise filtration; color segmentation; image retrieval; image fusion; dimensionality reduction for objects classification; search enhancement in large scale image databases, etc.
Linear and Non-linear Inverse Pyramidal Image Representation: Algorithms and Applications
Intelligent systems reference library, 2012
... Wavelets, CREW (Boliek et al.); ➢ Embedded block coding with optimized truncation of the embe... more ... Wavelets, CREW (Boliek et al.); ➢ Embedded block coding with optimized truncation of the embedded bit-streams, EBCOT (Taubman); ➢ Embedded Predictive ... Shiftable complex directional pyramid decomposition (Nguyen and Orain-tara); ❖ Improved multiresolution image ...
Extraction of human body contours and position analysis
2010 3rd International Congress on Image and Signal Processing, Oct 1, 2010
Abstract The paper presents a new method for extracting and positioning contours of moving object... more Abstract The paper presents a new method for extracting and positioning contours of moving objects. The method is applicable in the surveillance of elderly individuals and facilitates the detection of critical situations when the elderly individuals find themselves in need of ...
International Conference on Signal Processing, May 27, 2006
In the paper is presented a new approach for solving some of the authentication problems in large... more In the paper is presented a new approach for solving some of the authentication problems in large computer systems, communication networks and mobile communications, using a new method for lossless compression of some kinds of biometric information (fingerprints and signature images). The image processing is based on two-level Inverse Difference Pyramid (IDP) Decomposition with 2D Walsh-Hadamard Transform, followed by Histogram-Adaptive Run-Length data coding. In the paper are presented the comparison results, obtained for large number of test images of the pointed image classes. The investigation was performed with software products, based on the new method, on the JPEG 2000 standard (lossless version) and on the FBI compression standard. The new method attains high compression ratio and this is a basis for the future method development and implementation in security systems, which require fast and reliable user authentication. In the paper is presented a specific approach for digital watermarking based on the IDP and using the biometric data as a watermark.
Hierarchical Decomposition of Third-Order Tensor Through Adaptive Branched Inverse Difference Pyramid Based on 3D-WHT
Springer eBooks, 2022
In this work are presented some new approaches for efficient archiving of visual medical informat... more In this work are presented some new approaches for efficient archiving of visual medical information of various kinds. The main attention is aimed at the archiving of scanned paper documents. For this, new algorithms for image preprocessing and object segmentation are presented. The preprocessing is based on adaptive filtration, used to reduce the noises in the image background (corresponding to the image of the paper), retaining the main information (graphics, texts, etc.) untouched. The lossless compression, whose algorithm is given in detail, is based on new method for adaptive run-length coding, which comprises image histogram analysis and data coding. Significant advantage of the method is that it never allows enlargement of the coded files. The work comprises also experimental results, obtained using the software implementation of the algorithm and comparison with the well-known standards JPEG and JPEG 2000. The method is extremely efficient when used for coding of images of biomedical signals of any kind and has relatively low computational complexity, which is a reason to propose its use in telemedicine and in medical support appliances. Same approach is suitable for use in wide variety of applications (for example, telemedicine, etc.), which is proved by the experimental results included.
In the paper is presented one new method for multi-view object representation based on image deco... more In the paper is presented one new method for multi-view object representation based on image decomposition with Inverse Difference Pyramid. The method permits to obtain a very efficient description of the multi-view images using one of them as a reference one. The decomposition has a relatively low computational complexity because it is based on orthogonal transforms (Walsh-Hadamard, DCT, etc.). The relations which exist between transform coefficients from the consecutive decomposition layers permit significant reduction of the coefficients needed for the high-quality object representation.
Radix-(2×2) Hierarchical SVD for multi-dimensional images
The processing of multi-dimensional images is of high importance in contemporary communication an... more The processing of multi-dimensional images is of high importance in contemporary communication and information technologies. It requires significant computational resources, and is an object of large number of scientific investigations. One of the most efficient methods used, is the well-known Singular Value Decomposition (SVD), which permits the achievement of high compression together with significant reduction of the features' space, used for objects recognition. The main problem with the SVD is its high computational complexity. One approach to overcome the problem is presented in this paper. It is based on new decomposition for multi-dimensional images, which are treated as sequences of single high-correlated images through the so-called radix-(2×2) Hierarchical SVD (HSVD) algorithm. In correspondence with this approach, the multi-dimensional image is represented as a third-order tensor, divided into sub-tensors of size 2×2×2, called kernels. Each kernel is decomposed through Hierarchical SVD, based on the SVD for matrices of size 2×2, binary two-level tree and rearrangement of the components in each level. After kernel unfolding are obtained 2 matrices of size 2×2 and on each is applies SVD, calculated by using simple mathematical relations. In the paper is given a HSVD algorithm for a matrix of size 4×4, whose computational structure is described as a binary two-level tree. Same algorithm is used for the tensor decomposition of size 4×4×4. The decomposition is generalized for tensors of size N×N×N for N=2n>4. The computational complexity of the algorithm is evaluated and compared to that of the iterative SVD. The basic advantages of the new approach are the low computational complexity and the tree-like structure of the algorithm, which permits the low-energy leaves to be cut-off through threshold-based selection. As a result, the new algorithm is suitable for parallel processing of multi-dimensional images.
International Conference on Communications, Jul 13, 2006
In the paper is offered a new approach for blocking artifacts reduction in still images, processe... more In the paper is offered a new approach for blocking artifacts reduction in still images, processed with 2-level IDP Inverse Difference Pyramid (IDP) decomposition, based on DCT and Walsh-Hadamard orthogonal transforms. For this purpose was developed a two-dimensional fuzzy digital filter, whose performance changes in accordance with the image contents, framed by the filter window, and depending on the compression ratio and the maximum approximation error, obtained in result of the IDP decomposition. The experimental results show, that the block artifacts in the restored images are reduced without visual deterioration of the image sharpness. The main advantages of the new filter are its low computational complexity and the ability for adaptation in accordance with the image contents. In the paper are presented the results of the filter performance on JPEGcompressed images. The advantages of the new filter and its abilities for quality improvement of images with blocking artifacts are presented.
Lossless Image Compression with IDP and Adaptive RLC
CISST, 2004
Comparative Analysis of the Hierarchical 3D-SVD and Reduced Inverse Tensor Pyramid in Regard to Famous 3D Orthogonal Transforms
Springer eBooks, 2021
Truncated Hierarchical SVD for image sequences, represented as third order tensor
In this work is presented new algorithm, called Truncated Hierarchical SVD (THSVD), aimed at the ... more In this work is presented new algorithm, called Truncated Hierarchical SVD (THSVD), aimed at the processing of sequences of correlated images, represented as third-order tensors. The algorithm is based on the multiple calculation of the matrix SVD for elementary tensors (ET) of size 2×2×2, which build the tensor of size N×N×N, when N=2n. The new approach is compared to closest famous hierarchical SVD methods for ET: the Sequential Unfolding SVD (SUSVD) and the Radix 2×2Hierarchical SVD (Radix 2×2 HSVD). New two-level algorithm is developed for ET decomposition, with lower computational complexity than these of Radix 2×2 HSVD and SUSVD. In the paper is presented the THSVD algorithm for tensor of size 4×4×4, which is generalized for a tensor of size N×N×N. Adaptive new algorithm is offered for the “truncation” of the tensor decomposition components with small weights. The multiple execution of similar operations for the SVD calculation for matrices of size 2×2 in each THSVD level, permits its parallel implementation by using processors with relatively simple structures. As a result of the „truncation“ and of the parallel calculations of THSVD, the processing of image sequences represented by third-order tensors, is significantly accelerated. This advantage of the algorithm opens new abilities for its application in real-time image processing systems in various areas: compression of image sequences, digital watermarking, computer vision, machine learning, processing of multidimensional signals, etc.
Sliding Recursive Hierarchical Adaptive PCA for 3D image processing
Principal Components Analysis (PCA) is the basic approach for processing of 3D tensor images (for... more Principal Components Analysis (PCA) is the basic approach for processing of 3D tensor images (for example, multi- and hyper-spectral, multi-view, computer tomography, video, etc.). As a result of their processing, the information redundancy is significantly reduced. This is of high importance for their efficient compression and for the reduction of the features space needed, when object recognition is performed. The basic obstacle for the wide application of PCA, is the high computational complexity. One of the approaches, used to overcome the problem, is to use algorithms, based on the recursive PCA. The well-known methods for recursive PCA are aimed at the processing of sequences of images, represented as non-overlapping groups of vectors. In the last several years, the interest towards the recursive PCA for 3D tensor images executed in sliding temporal window, was significantly increased. Such analysis is needed, for example for the convolution tensor decomposition in multi-layer neural networks, in the systems for 3D noise filtration, video compression, etc. In this work is proposed new method, called Sliding Recursive Hierarchical Adaptive PCA (SR-HAPCA), based on the HAPCA algorithm. The method SR-HAPCA retains all advantages of HAPCA towards the PCA: it decreases the number of calculations needed, and permits parallel implementation. Besides, the lower computational complexity of SR-HAPCA makes easier its application in real-time processing of 3D tensor images.
Uploads
Papers by Roumiana Kountcheva