Andrei A. Kurekin

Data fusion and processing for airborne multichannel system of radar remote sensing: methodology, stages, and algorithms

Methodology and stages of data processing in multichannel airborne radar imaging systems are considered. It is shown that data fusion in such systems requires special techniques, algorithms, and software for image processing and information retrieval. Some approaches and methods are proposed. The results are demonstrated for simulated and real images.

Adaptive-vector LQ filter for color image processing

Robust adaptive vector filtering algorithms applicable to color and multichannel image processing are proposed. They are based on the use of Q-parameter that is a vector analog of quasirange. Considered algorithms have a good combination of properties: effective noise reduction, ability to remove spikes, edge and detail preservation, and low computational complexity. Their characteristics are evaluated quantitatively and compared to non-adaptive counterparts. Advantages of proposed algorithms are also demonstrated by simulated image processing results.

Soil moisture and erosion degree estimation from multichannel microwave remote sensing data

The techniques of soil moisture and erosion state estimation using multichannel airborne radar remote sensing system are considered. The experimental data proving the efficiency of the proposed approach and showing the dependence of scattered signal intensity on soil characteristics are presented. The algorithms of radar image preprocessing are discussed.

Soil-erosion-state interpretation using pre- and postprocessing of multichannel radar images

The multistage procedure of multichannel image pre- and post- processing is proposed. It includes nonlinear image-to-image and image-to-topology map superimposing with optimization of transform parameters, image separate and vector filtering for remained error reduction and image enhancement. Then the image preliminary recognition is to be performed for determined homogeneous regions, two possible ways of data interpretation based on the use of radiophysical models or supervised learning neural networks are discussed. It is shown that the proposed procedure provides a satisfactory determination of field lots with different erosion and other useful results.

Adaptive nonlinear vector filtering of multichannel radar images

Here we consider the possibility of applying vector median filtering for joint processing of multichannel radar images. The goal is to correct distortions due errors in superimposing of channels. The proposed approach is a two-stage one. It takes into account the difference between statistical characteristics of images formed by side-look aperture radar (SLAR) and synthetic aperture radar (SAR) and uses the correlation properties of the signal processed by vector filter. Image enhancement is achieved as many false edges are eliminated and contrast of true edges is increased. Filter properties are analyzed both with simulated test images and real radar data.

Modified vector sigma filter and its application to color and multichannel remote sensing radar image processing

A novel algorithm based on the sigma filter for processing multicomponent images is proposed. The noise suppression ability of the proposed vector filtering algorithm is better than, e.g., that of the standard sigma filter. Moreover, the added modifications make the filter able to remove impulsive noise. The proposed vector filter takes into account the mutual correlation between image components and preserves object edges and fine details even when the contrasts of the component images of multichannel data are low. The comparative analysis of filter performance is done both visually and using several quantitative criteria. Both simulated and real color and multichannel radar images are studied. It is shown that the modified vector sigma filter outperforms many component and vector filters. Two modifications are considered -- for cases of additive and multiplicative noise. Examples of the filter performance for processing real images formed by multipolarization/multifrequency side-look aperture radars are presented.

Processing multichannel radar images by modified vector sigma filter for soil erosion degree determination

A novel vector filter called modified vector sigma filter is proposed for processing the multichannel remote sensing radar images. It is demonstrated through simulations and real data examples that the proposed filter is able to provide an excellent combination of properties. It possesses efficient suppression of multiplicative noise and good edge preservation. Moreover, it simultaneously ensures an ability to remove spikes from images and excellent preservation of fine details even if they are characterized by rather low contrasts. These features occur to be useful for further interpretation of multichannel radar images, e.g. for determination of bare soil characteristics like erosion state. For simulated images it is shown that the application of the modified vector sigma filter is preferable in respect with its componentwise counterpart as the former technique provides less misclassification errors.

Application of order statistic filtering to multichannel radar image processing

The problem of secondary processing digital algorithm design and selection for multichannel remote sensing, radar images filtering, and first stage segmentation/recognition is discussed. On the basis of radar image specific features analysis it is shown that the application of order statistic algorithms is expedient for different stages of data processing. Some novel techniques and approaches to image filtering and interpreting are proposed.

Neural network application for primary local recognition and nonlinear adaptive filtering of images

An approach to neural network (NN) application to image local recognition based on several statistical parameter evaluations in a scanning window and their further joint analysis is put forward. Ways to deal with the images corrupted by dominant multiplicative or additive noise are discussed. The NN learning and structure selection methodologies are considered. The neural network classifier performance is analysed for the training and verification data sets. The possible applications of primary image recognition results, in particular, for further nonlinear locally adaptive filtering of images are proposed. The corresponding numerical simulation results proving the efficiency of considered techniques are proposed.

Radar dual-polarization remote sensing of soil erosion

The results of experiments for soil erosion determination with the use of dual-polarization radar set at wavelengths of 3 and 0.8 cm are presented. It is shown that the specific radar cross-section (RCS) is a function of soil erosion at the range of incidence angles from 35 degrees to 60 degrees. The sensitivity of soil backscattering to erosion is more high when the ratio of RCS for HH and VV polarizations is used. The best description of surface roughness spatial spectra is fractal one. The results of image processing obtained by airborne radar remote sensing system of Ku-band with dual- polarization reception are discussed and the comparison of results of soil erosion determination by radar remote sensing techniques with in situ measurements is made.