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

From the Publisher: 
The accessible presentation of this book gives both a general view of the entire computer vision enterprise and also offers sufficient detail to be able to build useful applications. Users learn techniques that have proven to be useful by first-hand experience and a wide range of mathematical methods. A CD-ROM with every copy of the text contains source code for programming practice, color images, and illustrative movies. Comprehensive and up-to-date, this book includes essential topics that either reflect practical significance or are of theoretical importance. Topics are discussed in substantial and increasing depth. Application surveys describe numerous important application areas such as image based rendering and digital libraries. Many important algorithms broken down and illustrated in pseudo code. Appropriate for use by engineers as a comprehensive reference to the computer vision enterprise.

Computer vision : a modern approach = 计算机视觉 : 一种现代的方法

.............................................................................................................................................................. 6 1 Introduction ............................................................................................................................................. 7 1.1 Motivation .......................................................................................................................................................... 7 1.2 Previous Surveys ............................................................................................................................................. 8 1.3 Overview of Technologies ........................................................................................................................... 9 1.4 Indoor Positioning Applications ............................................................................................................ 11 1.5 Structure of this Work ............................................................................................................................... 14 2 User Requirements ............................................................................................................................. 15 2.1 Requirements Parameters Overview .................................................................................................. 15 2.2 Positioning Requirements Parameters Definition ......................................................................... 17 2.3 Man Machine Interface Requirements ................................................................................................ 19 2.4 Security and Privacy Requirements ..................................................................................................... 20 2.5 Costs .................................................................................................................................................................. 20 2.6 Generic Derivation of User Requirements......................................................................................... 20 2.7 Requirements for Selected Indoor Applications ............................................................................. 21 3 Definition of Terms ............................................................................................................................. 25 3.1 Disambiguation of Terms for Positioning .......................................................................................... 25 3.2 Definition of Technical Terms ................................................................................................................ 27 3.3 The Basic Measuring Principles ............................................................................................................. 29 3.4 Positioning Methods ................................................................................................................................... 31 4 Cameras .................................................................................................................................................. 34 4.1 Reference from 3D Building Models .................................................................................................... 35 4.2 Reference from Images .............................................................................................................................. 36 4.3 Reference from Deployed Coded Targets .......................................................................................... 37 4.4 Reference from Projected Targets ........................................................................................................ 38 4.5 Systems without Reference ..................................................................................................................... 39 4.6 Reference from Other Sensors ............................................................................................................... 40 4.7 Summary on Camera Based Indoor Positioning Systems ........................................................... 40 5 Infrared ................................................................................................................................................... 42 5.1 Active Beacons .............................................................................................................................................. 42 5.2 Imaging of Natural Infrared Radiation ............................................................................................... 43 5.3 Imaging of Artificial Infrared Light ....................................................................................................... 43 5.4 Summary on Infrared Indoor Positioning Systems ....................................................................... 44

Indoor Positioning Technologies

The theory and practice of unmanned aerial vehicle UAV capture and control via Global Positioning System GPS signal spoofing are analyzed and demonstrated. The goal of this work is to explore UAV vulnerability to deceptive GPS signals. Specifically, this paper 1 establishes the necessary conditions for UAV capture via GPS spoofing, and 2 explores the spoofer's range of possible post-capture control over the UAV. A UAV is considered captured when a spoofer gains the ability to eventually specify the UAV's position and velocity estimates. During post-capture control, the spoofer manipulates the true state of the UAV, potentially resulting in the UAV flying far from its flight plan without raising alarms. Both overt and covert spoofing strategies are considered, as distinguished by the spoofer's attempts to evade detection by the target GPS receiver and by the target navigation system's state estimator, which is presumed to have access to non-GPS navigation sensor data. GPS receiver tracking loops are analyzed and tested to assess the spoofer's capability for covert capture of a mobile target. The coupled dynamics of a UAV and spoofer are analyzed and simulated to explore practical post-capture control scenarios. A field test demonstrates capture and rudimentary control of a rotorcraft UAV, which results in unrecoverable navigation errors that cause the UAV to crash.

Unmanned Aircraft Capture and Control Via GPS Spoofing

This newly revised and greatly expanded edition of the popular Artech House book Principles of GNSS, Inertial, and Multisensor Integrated Navigation Systems offers you a current and comprehensive understanding of satellite navigation, inertial navigation, terrestrial radio navigation, dead reckoning, and environmental feature matching . It provides both an introduction to navigation systems and an in-depth treatment of INS/GNSS and multisensor integration. The second edition offers a wealth of added and updated material, including a brand new chapter on the principles of radio positioning and a chapter devoted to important applications in the field. Other updates include expanded treatments of map matching, image-based navigation, attitude determination, acoustic positioning, pedestrian navigation, advanced GNSS techniques, and several terrestrial and short-range radio positioning technologies. The book shows you how satellite, inertial, and other navigation technologies work, and focuses on processing chains and error sources. In addition, you get a clear introduction to coordinate frames, multi-frame kinematics, Earth models, gravity, Kalman filtering, and nonlinear filtering. Providing solutions to common integration problems, the book describes and compares different integration architectures, and explains how to model different error sources. You get a broad and penetrating overview of current technology and are brought up to speed with the latest developments in the field, including context-dependent and cooperative positioning. DVD Included: Features eleven appendices, interactive worked examples, basic GNSS and INS MATLAB simulation software, and problems and exercises to help you master the material.

Principles of GNSS, Inertial, and Multisensor Integrated Navigation Systems, Second Edition

During the last years, there is an increasing demand for cheap and easy to operate platforms for surveillance and reconnaissance purposes. Therefore, the development of micro aerial vehicles is receiving an increasing attention. However, VTOL-MAVs often show an inherent instability that makes at least an automatic stabilization necessary, because otherwise the operator would not be able to keep these vehicles airborne. This requires the availability of navigation information, especially the vehicle's attitude has to be known. This paper addresses the development of an integrated navigation system based on MEMS inertial sensors and GPS for a VTOL-MAV. Special attention is paid to the handling of GPS outages. While usually periods without GPS aiding can be bridged using the unaided strapdown solution, the poor quality of the MEMS inertial sensors prohibits this approach here. Therefore, during GPS outages the accelerometer data is interpreted as approximate measurements of the local gravity vector. Additionally, the usage of a magnetometer providing measurements of the Earth's magnetic field is motivated and discussed. Finally, flight test results illustrate the performance of the resulting system, proving that the achieved attitude accuracy is sufficient for the automatic control of the MAV. This holds in situations with permanent GPS loss and dynamic maneuvering, too.

An Integrated GPS/MEMS-IMU Navigation System for an Autonomous Helicopter

An integrated GPS/MEMS-IMU navigation system for an autonomous helicopter

Tightly coupled GPS/INS integration for missile applications

In this paper, the fusion of GPS pseudorange and deltarange measurements with inertial sensor data is adressed. For many years, extended Kalman filters (EKF) have been applied for this task with great success. However, from a theoretical point of view, the EKF is a sub-optimal choice: The system dynamics model, which is given by the inertial navigation strapdown equations, as well as the pseudorange and deltarange measurement models are nonlinear. The EKF approximates the propagation of Gaussian random vectors through these nonlinear equations by a linear transformation. This allows to capture the variance-covariance matrix of the propagated Gaussian random vectors with first order accuracy only. The family of sigma-point Kalman filters (SPKF) offers an approximation of variance-covariance matrix which is accurate to at least second order. Therefore, the performance of EKF-based and SPKFbased tightly coupled GPS/INS systems is compared in numerical simulations. Different inertial sensor grades from MEMS to FOG and a variety of scenarios are investigated, including situations with less than four satellites in view. Additionally, the simulation results were confirmed by the post-processing of raw GPS and inertial sensor data that was recorded during a test drive. It was found that except for specific situations without practical relevance, EKF and SPKF offer an identical performance. This is due to the fact that for tightly coupled - as well as loosely coupled - GPS/INS integration the higher-order transformation terms are negligible, which is shown analytically.

A Performance Comparison of Tightly Coupled GPS/INS Navigation Systems Based on Extended and Sigma Point Kalman Filters

In this paper an Indoor Navigation System with map-matching capabilities in real-time on a smart phone is presented. The basis of the system is an in-house development of an Integrated Pedestrian Navigation System, based on 2 low-cost IMUs, an electronic compass and an altimeter with a drifting navigation solution. Combining this system with an additional laser ranger and SLAM algorithms, we are able to build accurate maps of office buildings for already visited rooms in post processing. This paper presents a map matching algorithm based on a new reduced particle filter in order to use these maps later for real-time applications without an expensive laser ranger but relying only on the dual inertial system. It can be used with both, preprocessed SLAM maps or with already available maps. Finally to smooth the resulting trajectory after particle filtering we propose the use of a new “balanced bubble band smoother” allowing the trajectory to optimally match to both, map and recorded IMU data. This new approach makes it possible to do map matching online on a smart phone.

Gert F. Trommer

Papers

An Integrated GPS/MEMS-IMU Navigation System for an Autonomous Helicopter

An integrated GPS/MEMS-IMU navigation system for an autonomous helicopter

Tightly coupled GPS/INS integration for missile applications

A Performance Comparison of Tightly Coupled GPS/INS Navigation Systems Based on Extended and Sigma Point Kalman Filters

Dual IMU Indoor Navigation with particle filter based map-matching on a smartphone