Vol 28, No 3 (2018)

We analyze the mathematical aspects of the data analysis problem consisting in the search (selection) for a subset of similar elements in a group of objects. The problem arises, in particular, in connection with the analysis of data in the form of time series (discrete signals). One of the problems in modeling this challenge is considered, namely, the problem of searching in a finite sequence of points from the Euclidean space for the subsequence with the greatest number of terms such that the quadratic spread of the elements of this subsequence with respect to its unknown centroid does not exceed a given percentage of the quadratic spread of elements of the input sequence with respect to its centroid. It is shown that the problem is strongly NP-hard. A polynomial-time approximation algorithm is proposed. This algorithm either establishes that the problem has no solution or finds a 1/2-approximate solution if the length M^* of the optimal subsequence is even, or it yields a \(\frac{1}{2}\left(\begin{array}{c}1-\frac{1}{M^*}\\ \end{array}\right)\)-approximate solution if M* is odd. The time complexity of the algorithm is O(N³(N²+q)), where N is the number of points in the input sequence and q is the space dimension. The results of numerical simulation that demonstrate the effectiveness of the algorithm are presented.

Discrete Tchebichef transform (DTT) has been utilized to improve the reconstruction quality of the traditional methods in image compression. Although DTT has the effective capability of energy concentration and ease of computation, not been exploited polynomials in orthogonal transform as compared with discrete cosine transform (DCT). This paper proposes an efficient lossy compression based DTT to produce better quality reconstructed image for the desired compression ratio. We combine soft decision quantization (SDQ) to design optimal quantization table and to approximate the rate-distortion for the purpose of the reconstruction quality. Compared with DCT under the scheme of JPEG baseline system, experimental results show that the proposed algorithm is of greater reconstruction image quality when the bit ratio exceeds 0.5 bpp. The bit ratio is decreased by 0.25, 0.49, 0.20 bpp, respectively when peak signal-to-noise-ratio (PSNR) is 35, 40, 45 dB. Meanwhile, they are similar on the elapsed time in encoding and decoding.

In last years, various medical image fusion algorithms have been proposed to fuse medical image. But, most of them focus on fusing grayscale images. This paper proposes a qualified algorithm for the fusion of multimodal color medical images. The technique of F-transforms has mainly been employed as a fusion technique for images obtained from equal or different modalities. The restriction of fused color mixing RGB, substitution method is resolved by incorporating F-transform and color mixing RGB. The proposed method significantly outperforms the traditional methods in terms of both visual quality and objective evaluation, with improved contrast and overall intensity. The proposed method provides better visual information than the gray ones and more adaptable to human vision. Additional, PCA is functional on the two-level decomposition to maximize the spatial resolution. Experimental evaluation demonstrates that the proposed algorithm qualitatively outperforms many existing state-of-the-art multimodal image fusion algorithms.

Nowadays, finding and Tracking a person in the world of technology is becoming a necessary task for various security purposes. Since the advent of technology, the development in the field of Facial Recognition plays an important role and has been exponentially increasing in today’s world. In this, a model is proposed for facial recognition to identify and alert the system when a person in search has been found at a specific location under the surveillance of a CCTV camera. The CCTV cameras are connected to a centralized server to which the live streaming feed is uploaded by cameras at each location. The server contains a database of all persons to be found. Based on the video feed from each camera, if a particular person in search is found in a certain feed, then the location of that person will be tracked and also a signal is passed to the system responsible. This model is based on image processing concepts to match live images with the existing trained images of the person in search. Since this model recognizes a person based on the first and foremost primary unique feature of a human, that is, only the person’s face image is required and will be found to be stored in the database. Hence the task of finding a person reduces to the task of detecting human faces in the video feed and matching with the existing images from the database.

Factors influencing the efficiency of biometrical personal identification based on cardiograms are studied. We compare the efficiency of cardiogram identification for healthy respondents and for respondents with various dysfunctions in the cardiovascular system. We study how sample size and time of cardiogram recording influence the effectiveness of recognition. Different methods of pattern recognition used in machine learning are compared. We also examine how different methods for cardiogram recording influence recognition accuracy.

Finger vein recognition biometric system is one of the most current and accurate biometric technologies. Yet, early implementation of this technology is not widely used in real-time applications. In this work, a finger vein-embedded system based on Rasberry-Pi has been presented. In our process, we use four structural directional elements for smoothing finger veins ROIs. A Top-Hat and Bottom-Hat kernel filters are used to enhance the contrast quality of images. For feature extraction step, we used two approaches for the synthesis of attributes including the geometric and texture representations of venous prints. The first one is a Local Directional Code (LDC) descriptor that characterizes texture and directional information of finger vein print. The Improved Gaussian Matched Filter (IMPGMF) is used to extract the finger vein map that characterised geometric venous information. The proposed vision system presents an Error Equal Rate (EER) lower to 0.02 and Identification Rate (IR) higher to 98.99. Moreover, experimental results show that the designed system is fast enough to run the decision of finger vein verification. Performance results show the efficiency and robustness of our system.

Precipitation is an important part of hydrological and meteorological models. For this reason, the development of adequate mathematical techniques and the design of software tools for the processing of large volumes of collected observations are important tasks. In particular, this refers to methods using modern approaches based on neural networks. In addition, studies of various precipitation processes are actual in the context of global warming and climate change. The paper is devoted to a detailed study of the possibility of constructing high-precision precipitation forecasts based on neural networks within patterns as a data mining technique for the meteorological data processing. A sufficiently high accuracy of forecasts is demonstrated for various characteristics of test patterns: up to 97% of one-day forecasts and up to 90% of two-day forecasts are successful. In the software sense, the work with neural networks is based on the deep learning library Keras for the programming language Python. For the sake of illustration, graphics are prepared using MATLAB software solutions.

Examining locomotion has improved our basic understanding of motor control and aided in treating motor impairment. Mice and rats are the model system of choice for basic neuroscience studies of human disease. High frame rates are needed to quantify the kinematics of running rodents, due to their high stride frequency. Manual tracking, especially for multiple body landmarks, becomes extremely time-consuming. To overcome these limitations, we proposed the use of superpixels based image segmentation as superpixels utilized both spatial and color information for segmentation. We segmented some parts of the body and tested the success of segmentation as a function of color space and SLIC segment size. We used a simple merging function to connect the segmented regions considered as a neighbor and having the same intensity value range. In addition, 28 features were extracted, and t-SNE was used to demonstrate how much the methods are capable to differentiate the regions. Finally, we compared the segmented regions to a manually outlined region. The results showed for segmentation, using the RGB image was slightly better compared to the hue channel. For merging and classification, however, the hue representation was better as it captures the relevant color information in a single channel.

Subject of Research. Application of barcoding technologies in the tasks of facial biometrics is posed and discussed. We analyze the achievements and estimate the shortcomings of existing solutions and examples of barcode creation according to the face images and the features extracted by them. Method. The ways of the problem implementation are determined and new solutions are presented based on linear (Code 128) and two-dimensional (QR) barcodes, as well as their color variants. The composition and volume of data are considered being used in the facial biometry and related applications: medicine, criminalistics and forensic-medical examination. Among these data there are face images, as well as sets of anthropometric points and additional information to them, information about the phenotype of face images (FI) and gender, and, finally, documentary information. Main Results. We have shown the results of these data “recording and transferring” within the framework of various barcode layouts, as well as the results of their reading and ways of hiding from reading. The proposed color barcodes are defined as “BIO Code 128” and “BIO QR-code”. While graphical display and computer memory record, they may be viewed as, colored raster images that carry information about the face in each layer. At this, documentary information may be read directly from such color images by standard barcode scanners, and the rest of the information (face image itself, its anthropometric, accompanying parameters) is read and restored after their decomposition into layers R, green and blue. Practical Relevance. The layout variants of the “BIO Code 128” and “BIO QR-code” barcodes and the programs for their generation (written in the MATLAB package environment) may be used in the further studies of the barcoding problem in the tasks of the facial biometrics and its applications.

In spite of a fact that many standalone iris recognition solutions are successfully implemented and deployed around the world, development of a reliable iris recognition solution capable to provide high recognition performance (both in biometric quality and speed) on mobile device is still an actual task. Main issues related to iris recognition in the mobile devices consist in uncontrollable capturing conditions and limitations in computation power. The aim of the proposed approach is to eliminate aforementioned issues by providing user with comprehensive feedback and, at the same time, performing the most computationally complex operations only on the images of the best quality. Key features of the proposed approach are multi-stage algorithm structure, novel iris image quality estimation and adaptive iris feature vector quantization algorithms. These features allow to achieve high recognition accuracy and real-time performance which are proved by experimental results.

Watermarking is the advanced technology to illuminate the issue of securing sight and secret data. It is the ability of covering the data into a host with the end goal that the inserted information is undetectable. In this proposed approach layer fusion based technique is employed, alternative pixels from the chrominance layers forms a merged layer and mark the transpose value of the distinctive pieces of color copyright information under wavelet. Finally, the layers retain the original position form a watermarked video frame. The experimental results on video sequences are showing the effectiveness better than the existing techniques in terms of Peak Signal to Noise Ratio (PSNR) and Correlation Coefficient (CC). Likewise, the results indicate that the framework can be strong against regular video processing attacks.

In the wavelet transform domain, the wavelet thresholding denoising is an effective noise reduction method for noisy images. The key issues of image thresholding denoising are the choice of threshold value and construction of thresholding function. To overcome the shortcomings in the classical wavelet thresholding methods such as fixed threshold value and inflexible thresholding function, a modified artificial bee colony (MABC) algorithm-based parametric wavelet thresholding approach is utilized. A construction scheme of parametric wavelet thresholding function is firstly put forward. And a new tangent function-based thresholding is built based on the construction method. The threshold value and shape tuning parameter of the proposed thresholding are initialized as the possible solutions of the MABC algorithm, and the corresponding objective function is minimized. Finally, the MABC-based approach is applied to process two types of images with different degrees of degradation. It is also compared with classical thresholding and other optimization algorithm-based methods in terms of different criteria. Comparison results demonstrate that the proposed MABC-based thresholding approach achieves better enhancement in terms of denoising capability.