Timofei V. Kondranin

Retrieval of forest stand attributes using optical airborne remote sensing data

Optical remote sensing data processing is proposed for the airborne images of high spectral and spatial resolution. Optimization techniques are undertaken to gain information about spatial distribution of the pixels on the hyperspectral images and the texture of the forest stands of different species and ages together with reducing redundancy of the spectral channels used. The category of neighborhood of pixels for particular forest classes and the step up method of selecting optimal spectral channels are employed in the relevant processing procedures. We present examples of pattern recognition for the forests as a result of separating pixels, which characterize the sunlit tops, shaded space and intermediate cases of the Sun illumination conditions on the hyperspectral images.

Effectiveness of the polarization discrimination technique for underwater viewing systems

Limits of the polarization technique in discriminating backscattered noise for a monostatic underwater viewing system are discussed. The approach presented is based on an analytical solution of the vector equation for radiative transfer in the small angle approximation. The influence of the water optical properties, the receiver range gate, and the field of view on the background reduction are investigated. Estimates are made of the enhancement of the targets apparent contrast and of the detection range that can be achieved for typical environmental conditions.

Recognition of natural and man-made objects in airborne hyperspectral images

New approaches to the processing of airborne hyperspectral images are implemented in order to develop emerging applications based on high-performance computing resources. The focus is on solving the problem of recognizing forest vegetation of different species composition and age based on high spectral and spatial resolution airborne sensing data. Examples of the formation of information layers of recorded spectra for “pure species” of pine and birch forests are given with the selection of illuminated and shaded pixels, which increases the accuracy of recognition of objects in the processing of these images.

Optimization of the polarization remote-sensing techniques of the ocean

A numerical code has been developed to calculate Stokes parameters of the visible solar radiation, scattered in the atmosphere-ocean system. Mathematical modeling is used to examine spectral and angular (azimuth and zenith angle) variations of degree of polarization at sea level and at different heights in the atmosphere above the sea surface. On the basis of a developed computer code the efficiency of the polarization measurements for different optical passive remote sensing techniques of the ocean has been investigated. For the passive spectral measurements of the water bio-productivity (chlorophyll-a, dissolved organic matter, concentration of suspended particles) the polarizer can improve signal-to-background ratio. The magnitude of this effect and optimum direction of the polarizer depend upon height, viewing direction, and solar zenith angle. Within the framework of polarization remote sensing technique the influence of the observation height and viewing direction on the results of water turbidity measurements is investigated. Optimal viewing directions in such polarization passive remote sensing technique are discussed.

Comparison Analysis of Recognition Algorithms of Forest-Cover Objects on Hyperspectral Air-Borne and Space-Borne Images

The basic model for the recognition of natural and anthropogenic objects using their spectral and textural features is described in the problem of hyperspectral air-borne and space-borne imagery processing. The model is based on improvements of the Bayesian classifier that is a computational procedure of statistical decision making in machine-learning methods of pattern recognition. The principal component method is implemented to decompose the hyperspectral measurements on the basis of empirical orthogonal functions. Application examples are shown of various modifications of the Bayesian classifier and Support Vector Machine method. Examples are provided of comparing these classifiers and a metrical classifier that operates on finding the minimal Euclidean distance between different points and sets in the multidimensional feature space. A comparison is also carried out with the “K-weighted neighbors” method that is close to the nonparametric Bayesian classifier.

Multichannel aerospace multispectral and hyperspectral imagery registration system designed for remote sensing and investigation of the earth surface, ocean, and atmosphere

The paper outlines an aerospace system intended for registration of multispectral and hyperspectral imagery within 250 - 1000 nm. The system is a versatile tool for remote sensing and investigation of the Earth surface, ocean and atmosphere. It includes: medium-resolution multispectral linear CCD array-based device providing simultaneous viewing in 3 of 8 relatively narrow spectral subranges within 400 - 1000 nm band; low-resolution hyperspectral CCD array-based image spectrometer providing simultaneously up to 512 spectral counts within 250 - 1000 nm band. The system makes it possible to set resolution and spectral parameters to fit the needs of various application tasks and observation conditions.

Electro-optical modules for space monitoring of transportation systems

Electro-optical modules based on multispectral photodetectors are used for monitoring of Earths atmosphere and surface. This monitoring is highly effective and allows to provide the ecological control of environmental pollution and possible aftereffects of such pollution. This paper analyzes the possibility of photonic system application for ecological monitoring and emergency control of railroad transportation.

Estimation of the uncertainties of the retrieved seawater parameters

A method is proposed to estimate the accuracy of parameters retrieval from the spectral sea reflectance based on certain assumptions about errors distribution in the input data. Maximum likelihood method has been used to estimate the error of the retrieval stemmed from the random errors of experimental data. The method has been applied to the retrieval of chlorophyll-a concentration, coefficient of scattering in the backward direction of non-chlorophylous particles and absorption of yellow substance in the waters of north-east part of the Black sea. The retrieved chlorophyll concentrations are consistent with in situ measurements. The modeled values of scattering coefficient check well with measurements of extinction coefficient. Proposed accuracy estimates proved to be close to direct ones. The proposed method is general enough to be applicable to the accuracy estimation in various inverse problems of remote sensing of the ocean and atmosphere.