Digital Signal Processing

Russian
Scientific & Technical
Journal


“Digital Signal Processing” No. 3-2013

Digital image processing

In the issue:

- video encoding standard HEVC;
- object coordinates estimation;
- reconstructions distorted signal of digital images;
- multidimensional EOF analysis in geoinformatics;
- geometric transformation;
- multispectral images;
- forming of color images;
- decoding algorithms for LDPC codes;
- DSP in optoelectronic devices.



N
EW POSSIBILITIES OF VIDEO ENCODING STANDARD HEVC
Dvorkovich V.P., Dvorkovich A.V., Gruzov G., dvp@niircom.ru

A detailed review of possibilities and algorithms included into new video encoding standard HEVC is presented. The description of standard and it characteristic features include target applications, main tasks of researches, general scheme of encoder and principal differences with previous standards. High level syntax of HEVC standard is derived from syntax of AVC standard but has a set of improvements and novelties. Differences of video encoding layer of HEVC with respect to AVC are analyzed. Finally a brief results of HEVC coding efficiency is presented showing growth in 35-40% using PSNR measure and more than 30-67% using MOS with respect to AVC coding.

Keywords: video encoding, standard, compression, parallel processing, HEVC, AVC, MPEG.

OBJECT COORDINATES ESTIMATION ALGORITHM BASED ON RADON TRANSFORM
Babayan P.V., Shubin N.J., aitu@rsreu.ru

The problem of large size artificial object coordinates estimation is described in this paper. The algorithm based on object outline estimation is suggested. It can be used when image transformation parameters are vastly changed from frame to frame. The results of experimental research of the proposed algorithm are presented.

Keywords: image processing, coordinate estimation, radon transform.

METHOD OF THE INTERCHANNEL GRADIENT RECONSTRUCTION OF DISTORTED SIGNALS OF DIGITAL COLOUR IMAGES
Samoylin E.A., Shipko V.V., es977@mail.ru, shipko.v@bk.ru

The reconstruction method of distorted or lost signal components of digital color images on local gradients in nearby channel is presented. Efficiency of this method is shown on the example of processing the distorted impulse noise of digital color images.

Keywords: color images, gradient reconstruction, model, applicative interference, filtering.

A
PPLICATION OF MULTIDIMENSIONAL EOF-ANALYSIS IN GEOINFORMATICS
Korotchenko R.A., Samchenko A.N., Yaroshchuk I.O.,
romankor@mail.ru


Principles of multidimension EOF-methods for processing and analyzing of geospatial data fields are considered. Procedural stages of two-dimensional singular spectrum analysis (2D-SSA) for spatial structural decomposition of the scalar fields are described.
For a part of the Sea of Japan north-western shelf a digital terrain model is developed and two-dimensional EOF-decomposition and further reconstruction of the model are carried out to select and to interpret spatial and temporal scales for observed geomorphological and tectonics structures.

Keywords:
multidimensional EOF-analysis, geoinformatics, two-dimensional spectrum analysis.

ALGORITHMS OF COLOR FUSION FROM MONOCHROME MULTISPECTRAL IMAGES BASED ON COLOR CONVERSION METHODS
Bogdanov A.P., Kholopov I.S., kholopov.i.s@rsreu.ru

The algorithms of digital monochrome visible and infrared image processing which form the color fusion image with the closer to daylight conditions color spectrum is considered. Computational capacity and the possibility of real-time implementing for transfer and replacing color algorithms in color space YCBCR is defined.


Keywords: color fusion, monochrome multispectral images, color conversion, image processing.

IMPROVE PERFORMANCE OF ALGORITHMS FOR FORMA COLOR FUSION IMAGE FROM MULTISPECTRAL MONOCHROME SENSORS SIGNALS
Bogdanov A.P., Pavlov O.V., Kholopov I.S., kholopov.i.s@rsreu.ru


Simplified algorithms of imaging in false colors in a decorrelated color space for the task of forming a color image with color range approximated daylight conditions is presented. It is shown that the performance of the proposed method of color transfer is up to 1.8 times higher compared with algorithms forming false color images in the space of RGB. An improvement and fusion algorithm of multispectral monochrome images is considered.

Keywords: color spaces, images aggregation, false colors, color transfer, bilateral filtering.

FORMING OF COLOR IMAGES COMPONENTS FROM THE DATA OF MULTISPECTRAL IMAGERY
Kuznetsov A.E., Pobaruev V.I., Svetelkin P.N., foton@rsreu.ru

The algorithms of forming multispectral image components of the Earth's surface from "Resurs-DK" and the "Meteor-M" satellites in natural colors is given in the paper. It is proposed to use a modified hyperspherical colorizing algorithm for obtaining high quality color images with high spatial resolution from panchromatic and multispectral images.

Keywords:
image processing, multispectral imagery, hyperspherical colorizing algorithm, panchromatic images.

R
EDUCED-COMPLEXITY BP-BASED DECODING ALGORITHMS FOR LDPC CODES
Likhobabin E.A.,
tor@rsreu.ru

Review of various reduced-complexity belief propagation (LLR-BP) based decoding algorithms for LDPC codes is presented. Complexity, effectiveness, and average number of iteration for each algorithm are shown.

Keywords: decoding algorithms, forward error correction, LDPC, reduced-complexity algorithms.

FEATURES AND ALGORITHMS OF DIGITAL SIGNAL PROCESSING IN OPTOELECTRONIC DEVICES
Molodyakov S.A., molod@cef.spbstu.ru


Digital signal processing in optoelectronic devices used to process optical signals from the output image sensors is considered in the paper. Algorithms for flow digital data processing are discussed as applied to radio astronomy.

Keywords: digital signal processing, optoelectronics, optical signal, image sensor, radio astronomy.

M
ETHOD OF CALIBRATING THE FREQUENCY SCALE PULSAR OPTOELECTRONIC PROCESSOR AND ITS IMPLEMENTATION USING A DSP
Lavrov A.P., Molodyakov S.A.,
molod@cef.spbstu.ru

Calibration method and equipment for digital signal processing in optoelectronic processor registration pulsed radio emission of cosmic sources - pulsars is described. Digital processing is carried out after receiving a signal from the output of the analog part of the processor. The calibration method involves the introduction of a new mode of the processor, and the measurement of the main parameters of the analog section in order to incorporate them into the profile of the resulting integral radio pulses. It is shown how to implement the method using reset program in the digital signal processor and a secondary processing.

Keywords: optoelectronic processor, digital signal processor, acousto-optic spectrum analyzer, frequency scale, drift, calibration, reload program, resampling.


If you have any question please write: info@dspa.ru