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

Aiming at airborne three-channel SAR-GMTI system and its real data, a novel channel equalization method is presented, furthermore, in virtue of the method's good performance on compensating the clutter differences between different channels, a technique for moving target detection (MTD) and location is further proposed. This method first realizes signal equalization between inner channels utilizing the leaking signals, then implements range and azimuth time calibration and compensates the channel differences via iteration after range compression, after that, it realizes MTD by clutter cancellation and imaging, finally obtains target parameters by estimating moving target doppler shift. The technique proved its good performance in real data processing with the advantages of high channel equalization precision, little computations, good clutter cancellation and MTD performances and easiness to implement.

A novel approach to three-channel SAR-GMTI channel equalization and moving target detection and location based on real data

The microwave photonic (MWP) radar has the capability of generating ultrawideband (UWB) signals. It is a challenge to realize accurate extraction of attributed scattering centers (ASCs) from MWP signals. This manuscript presents a scattering parameter estimation method in the image domain for UWB MWP signals. The polar-to-rectangular resampling is required for UWB MWP signals. Therefore, a range-azimuth decoupled representation based on the ASC model is formed. The model parameter estimation is converted into an optimization problem, where the statistics of the target signal and the features of interest are modeled to provide prior information. The distribution spread maximization (DSM) and peak magnitude maximization (PMM) principles in the optimization embody this prior information. The particle swarm optimization (PSO) is utilized to search for the parameters of each ASC in the image domain. Moreover, the orthogonal matching pursuit (OMP) algorithm is introduced to avoid repeated computation. Experimental results conducted on the simulated data, XPATCH data, and real data confirm the effectiveness of the proposed method. The proposed method takes into account the specific features of UWB MWP signals, which are neglected in the existing studies. Therefore, the proposed method performs better in extracting ASC parameters from UWB MWP signals in terms of accuracy and more complete sets.

Attributed Scattering Center Extraction Method for Microwave Photonic Signals Using DSM-PMM-Regularized Optimization

In this paper, a novel signal reconstructing method for radar targets is proposed based on the attributed scattering center model. By extracting the attributed parameters, the large amount of target data can be represented by small amounts of attributed parameters. In this way, the data amount has been compressed sharply, which releases the computer memory for storage. After extraction, a target discriminating method is presented by applying a CFAR threshold to the energy of extracted attributed scattering centers, by which, weak distributed scattering centers with relatively high energy in total can be discriminated from noise under low SNRs. Experimental results validate the effectiveness of the signal reconstructing capability of the proposal.

A novel signal reconstructing method for radar targets

A suitable regularization parameter plays an important role in sparse ISAR imaging algorithms. With a proper regularization parameter, the quality of ISAR images improves. In this paper, the Homotopy re-weighted l1-norm minimization is applied to ISAR imaging. This method is able to choose the accurate regularization parameter for each point in ISAR image with high efficiency. As a result, the imaging results processed by this method contain more details of the target and less artificial points. Both simulated and real data experiments validate the feasibility of the proposed method.

ISAR Imaging Based on Homotopy Re-Weighted ℓ1-Norm Minimization

In this article, we exploit a granular-based modeling scheme to realize the number of sources detection under nonideal conditions (in low signal-to-noise ratio levels and small snapshots scenarios), in which the idea of information granules and granular computing is integrated with the fuzzy set theory. In the developed scheme, a collection of eigenvalues, which is calculated by the array output correlation matrix, is constructed as a time series. Then, the principle of justifiable granularity criterion is introduced to split the time series into two regions so as to establish a set of upper and lower bounds of prototypes for the signal and noise subspaces. Subsequently, the fuzzy C-means-based encoding and decoding mechanism is employed to optimize the prototypes. During this process, the encoding mechanism is used to produce a pair of information granules, and the decoding mechanism is used to evaluate the quality of information granules. After several rounds of iteration, an optimized prototype vector and a partition matrix are generated. Finally, the structure of the time series is exposed (encoded into) by an optimized prototype vector and a partition matrix, and the number of sources can be determined conveniently through the partition matrix. Simulation results show that the proposed method outperforms the commonly used methods under nonideal conditions.

Mengdao Xing

Papers

A novel approach to three-channel SAR-GMTI channel equalization and moving target detection and location based on real data

Attributed Scattering Center Extraction Method for Microwave Photonic Signals Using DSM-PMM-Regularized Optimization

A novel signal reconstructing method for radar targets

ISAR Imaging Based on Homotopy Re-Weighted ℓ1-Norm Minimization

Modeling of Number of Sources Detection Under Nonideal Conditions Based on Fuzzy Information Granulation