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

The interpretation of Synthetic Aperture Radar (SAR) image is considered to be a challenging task, especially when tracking the target shadow in Video SAR (ViSAR), the speckle noise needs to be considered. Based on this, this paper proposes a SAR image denoising algorithm based on the improved wavelet threshold function. Different from the existing denoising methods, this algorithm combines the characteristics of hard threshold function and soft threshold function in traditional wavelet transform denoising, constructs a new threshold function, and improves the equivalent number of looks (ENL) of denoised SAR image. When the denoised image is applied to the tracking task, the target features are enhanced by k-means algorithm and binarization method, so as to improve the tracking accuracy. Experimental results show that the algorithm improves the tracking accuracy on the basis of ensuring the real-time performance of tracking and makes the tracking task highly robust to the noise of SAR image.

A SAR Image Denoising Method for Target Shadow Tracking Task

Refocusing moving targets in synthetic aperture radar (SAR) images is a challenging task Because of the unknown motion parameters of the targets. Thus, exact parameter estimation is required in the moving targets reconstructed. In order to solve the question, this paper proposed an optimization algorithm of moving targets refocusing via parameter estimation dependence of maximum sharpness principle after back projection (BP) integral. This method consists of three groups: Firstly, moving targets is detected and extracted from SAR images dependence of BP algorithm. Then, the extra phase brought by the motion parameters are obtained by driving exact function of target's 2-D wavenumber spectrum. Finally, based on the maximum sharpness principle, the motion parameters are optimized by iteratively compensating the extra phase. Moving targets can be focused well by removing the extra phase via the estimated parameters. Both simulation data and real data processing is used to demonstrate the effectiveness of the proposed algorithm.

An Optimization Algorithm of Moving Targets Refocusing Via Parameter Estimation Dependence of Maximum Sharpness Principle After BP Integral

Phase unwrapping (PU) is a key step for the synthetic aperture radar (SAR) interferometry (InSAR). Single-baseline (SB) PU and multi-baseline (MB) PU are two independently developed technologies, each of which has its own advantages and disadvantages. A two-stage programming-based MB PU method (TSPA) proposed by Yu [1] establishes a connection between the MB and SB PU methods. TSPA breaks the limitation of the phase continuity assumption by using the Chinese remainder theorem (CRT), and uses the minimum-cost flow (MCF) optimization model to obtain the PU result. TSPA can be regarded as a framework for solving MB PU problems. In this paper, we studied how to transplant the infinity-norm ( $L^{\infty}$ -norm) optimization model into TSPA framework. Under the TSPA MB PU framework, a $L^{\infty}$ -norm based MB PU method (referred to as Inf-TSPA) is proposed to solve the problem of low PU accuracy of the $L^{\infty}$ -norm SB PU method. The experimental results on the simulated and the realistic MB InSAR data sets verify that the performance of Inf-TSPA is significantly improved compared to the $L^{\infty}$ -norm SB PU method.

An Infinity-Norm-Based Phase Unwrapping Method with TSPA Framework for Multi-Baseline SAR Interferograms

Because fast factorized back-projection (FFBP) algorithm is not limited by the assumption of azimuth-invariant of echo signal, it has significant advantages for maneuvering platform synthetic aperture radar (MP-SAR) imaging. Due to the curved trajectory of MP-SAR, the focusing quality of the SAR image becomes very sensitive to the terrain scene elevation, and the range and angular histories are difficult to be solved in polar coordinate system for FFBP implementation. In this letter, a new image reconstruction algorithm integrated with elevation information is proposed for MP-SAR imaging with high efficiency. The proposed algorithm is based on the hybrid coordinate system which is incorporated with elevation information in the FFBP process. In the proposed algorithm, a flexible matching region is introduced to efficiently solve the range and angular histories in FFBP recursion which can reconstruct high-quality images of terrain scene with both high accuracy and efficiency. Simulation experiments are implemented and analyzed to validate the superior performance of the proposed algorithm.

An Efficient Image Reconstruction Algorithm for Maneuvering Platform SAR Integrated With Elevation Information in Hybrid Coordinate System

This letter mainly considers the environmental clutter problem in distinguishing between stationary humans and animals through-wall circumstances. Focusing on the challenges of object identification in the time–frequency map, we propose a cross-scale feature aggregation (CSFA) network based on channel–spatial attention, which can improve the identification accuracy of stationary humans and animals. Specifically, life detection radar is utilized to collect data, and the time–frequency analysis method synchrosqueezing transform (SST) is used to suppress the signal noise and generate higher-resolution time–frequency maps. In order to make full use of the target information, we use a feature pyramid network (FPN) to obtain multilevel feature information maps from time–frequency maps. Then, the CSFA module is utilized to extract detailed micro-Doppler feature information from feature maps. And we use a deep convolutional neural network (CNN) to classify humans from animals. Experimental results show that the proposed model has a better performance in accuracy compared with the existing methods.

Mengdao Xing

Papers

A SAR Image Denoising Method for Target Shadow Tracking Task

An Optimization Algorithm of Moving Targets Refocusing Via Parameter Estimation Dependence of Maximum Sharpness Principle After BP Integral

An Infinity-Norm-Based Phase Unwrapping Method with TSPA Framework for Multi-Baseline SAR Interferograms

An Efficient Image Reconstruction Algorithm for Maneuvering Platform SAR Integrated With Elevation Information in Hybrid Coordinate System

A Cross-Scale Feature Aggregation Network Based on Channel–Spatial Attention for Human and Animal Identification of Life Detection Radar