Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Обнаружение транспортных средств на изображениях загородных шоссе на основе метода Single shot multibox Detector

This chapter introduces the subject of statistical pattern recognition (SPR). It starts by considering how features are defined and emphasizes that the nearest neighbor algorithm achieves error rates comparable with those of an ideal Bayes’ classifier. The concepts of an optimal number of features, representativeness of the training data, and the need to avoid overfitting to the training data are stressed. The chapter shows that methods such as the support vector machine and artificial neural networks are subject to these same training limitations, although each has its advantages. For neural networks, the multilayer perceptron architecture and back-propagation algorithm are described. The chapter distinguishes between supervised and unsupervised learning, demonstrating the advantages of the latter and showing how methods such as clustering and principal components analysis fit into the SPR framework. The chapter also defines the receiver operating characteristic, which allows an optimum balance between false positives and false negatives to be achieved.

Statistical Pattern Recognition

Object detection in optical remote sensing images, being a fundamental but challenging problem in the field of aerial and satellite image analysis, plays an important role for a wide range of applications and is receiving significant attention in recent years. While enormous methods exist, a deep review of the literature concerning generic object detection is still lacking. This paper aims to provide a review of the recent progress in this field. Different from several previously published surveys that focus on a specific object class such as building and road, we concentrate on more generic object categories including, but are not limited to, road, building, tree, vehicle, ship, airport, urban-area. Covering about 270 publications we survey (1) template matching-based object detection methods, (2) knowledge-based object detection methods, (3) object-based image analysis (OBIA)-based object detection methods, (4) machine learning-based object detection methods, and (5) five publicly available datasets and three standard evaluation metrics. We also discuss the challenges of current studies and propose two promising research directions, namely deep learning-based feature representation and weakly supervised learning-based geospatial object detection. It is our hope that this survey will be beneficial for the researchers to have better understanding of this research field.

A survey on object detection in optical remote sensing images

In this paper, we examine the role of polarimetry in synthetic aperture radar (SAR) interferometry. We first propose a general formulation for vector wave interferometry that includes conventional scalar interferometry as a special case. Then, we show how polarimetric basis transformations can be introduced into SAR interferometry and applied to form interferograms between all possible linear combinations of polarization states. This allows us to reveal the strong polarization dependency of the interferometric coherence. We then solve the coherence optimization problem involving maximization of interferometric coherence and formulate a new coherent decomposition for polarimetric SAR interferometry that allows the separation of the effective phase centers of different scattering mechanisms. A simplified stochastic scattering model for an elevated forest canopy is introduced to demonstrate the effectiveness of the proposed algorithms. In this way, we demonstrate the importance of wave polarization for the physical interpretation of SAR interferograms. We investigate the potential of polarimetric SAR interferometry using results from the evaluation of fully polarimetric interferometric shuttle imaging radar (SIR)-C/X-SAR data collected during October 8-9, 1994, over the SE Baikal Lake Selenga delta region of Buriatia, Southeast Siberia, Russia.

Polarimetric SAR interferometry

Sparse frequency agile linear frequency modulation-orthogonal frequency division multiplexing (LFM-OFDM) radar can improve anti-interference ability as well as reduce the sampling rate of the radar system. However, the frequency agility makes it difficult to realize coherent processing by using the fast Fourier transform. To overcome this problem, a signal processing scheme for the sparse frequency agile LFM-OFDM radar is proposed to obtain the high range and Doppler resolution in this study. The subcarriers are synthesised to an LFM signal in the time domain at first. Then, within the framework of the compressed sensing theory, an improved orthogonal matching pursuit algorithm is proposed to obtain the high-resolution range-velocity profile while reducing the computation complexity. Finally, several numerical simulations are provided to verify the effectiveness of the proposed method for moving targets. 

Range and Doppler reconstruction for sparse frequency agile linear frequency modulation‐orthogonal frequency division multiplexing radar

With the constraint of the minimum antenna area in space-borne synthetic aperture radar (SAR), to obtain wide swath and high azimuth resolution, azimuth ambiguities are usually inevitable. With multiple channels data, beamforming can be utilized to remove the azimuth ambiguities. While in real multiple-channel data, there exist mismatch between different channels and different beams. A calibration method is proposed in this paper, which equalize the mismatch between channels and beams. Processing results of real multiple-channel data are presented to demonstrate its effectiveness.

Raw data based azimuth ambiguities removing

Ship oscillatory motions with long coherence processing interval (CPI) are complex and hard to accurately estimate. The traditional ship imaging methods usually avoid long-CPI focusing by dealing with a short observation time interval, which may make it hard to obtain a high-resolution and high-SNR image. In this paper, the mechanisms and challenges of long CPI imaging of oscillatory targets are investigated. The properties of wavenumber domain support (WDS) and point spreading function (PSF) of oscillatory targets are analyzed. It’s found that the WDS spreads as a three-dimensional (3D) thin and curved sweep surface, with time-variant energy density, non-parallel boundaries and a complex structure. The PSF of an oscillatory target has a 3D resolution, but also multiple side-lobes with high level. We inspect the relationships between the properties of the WDS and the non-ideal PSF. Moreover, it’s found that scatters distributed on a 3D oscillatory target cannot be focused uniformly on a predefined imaging plane (IP). The projection relationship of the target focusing positions on the slant-range plane (SRP) and the IP are also derived. The simulation results can well validate the proposed method.

Challenges of Ship Focusing with Long Coherence Processing Interval

In Airborne radar imaging of ship targets on the river (or the sea), itpsilas difficult to image each targets because each individual target may have its own motion characteristic different from others. Aim at this problem, this paper proposed a Airborne hybrid SAR/ISAR method imaging of muti-ship targets. By this method, the raw data is processed by SAR algorithm firstly, then, extract and recover of selected ship data from SAR image. Finally, ISAR image of selected targets. the effectiveness of the proposed method is verified by the real data processing.

A method for multiple ship target imaging based on hybrid SAR/ISAR method

Due to the effects of the earth's rotation and the satellite's elliptical orbit, the Doppler centroid varies along the orbit in geosynchronous earth orbit synthetic aperture radar (GEO SAR). With an ultrahigh orbit height, the beam may illuminate outside the earth surface with a rotation angle large than 9 degrees. Therefore, the usual attitude steering methods to minimize the Doppler centroid in low earth orbit SAR (LEO SAR) may not be suitable for GEO SAR. Considering the above two effects of GEO SAR and the beam illuminating restriction, a beam determination method to minimize Doppler centroid in GEO SAR is proposed in this paper. Guaranteeing that the beam not illuminate outside the earth surface, the proposed method can drastically decrease the Doppler centroid and the equivalent squint angle.

Mengdao Xing

Papers

Range and Doppler reconstruction for sparse frequency agile linear frequency modulation‐orthogonal frequency division multiplexing radar

Raw data based azimuth ambiguities removing

Challenges of Ship Focusing with Long Coherence Processing Interval

A method for multiple ship target imaging based on hybrid SAR/ISAR method

A novel beam direction determination method for minimizing Doppler centroid in GEO SAR