Log In
Or create an account ->
Imperial Library
Home
About
News
Upload
Forum
Help
Login/SignUp
Index
Cover
Title
Copyright
Dedication
Contents at a Glance
Contents
About the Authors
About the Technical Reviewer
Acknowledgments
Chapter 1: Introducing the Kinect
Hardware Requirements and Overview
Installing Drivers
Windows
Linux
Mac OS X
Testing Your Installation
Getting Help
Summary
Chapter 2: Hardware
Depth Sensing
RGB Camera
Kinect RGB Demo
Installation
Making a Calibration Target
Calibrating with RGB Demo
Tilting Head and Accelerometer
Summary
Chapter 3: Software
Exploring the Kinect Drivers
OpenNI
Microsoft Kinect SDK
OpenKinect
Installing OpenCV
Windows
Linux
Mac OS X
Installing the Point Cloud Library (PCL)
Windows
Linux
Mac OS X
Summary
Chapter 4: Computer Vision
Anatomy of an Image
Image Processing Basics
Simplifying Data
Noise and Blurring
Contriving Your Situation
Brightness Thresholding
Brightest Pixel Tracking
Comparing Images
Thresholding with a Tolerance
Background Subtraction
Frame Differencing
Combining Frame Differencing with Background Subtraction
Summary
Chapter 5: Gesture Recognition
What Is a Gesture?
Multitouch Detection
Acquiring the Camera Image, Storing the Background, and Subtracting
Applying the Threshold Filter
Identifying Connected Components
Assigning and Tracking Component IDs
Calculating Gestures
Creating a Minority Report—Style Interface
Considering Shape Gestures
Summary
Chapter 6: Voxelization
What Is a Voxel?
Why Voxelize Data?
Voxelizing Data
Manipulating Voxels
Clustering Voxels
Tracking People and Fitting a Rectangular Prism
Summary
Chapter 7: Point Clouds, Part 1
Representing Data in 3-D
Voxels
Mesh Models
Point Clouds
Creating a Point Cloud with PCL
Moving From Depth Map to Point Cloud
Coloring a Point Cloud
From Depth to Color Reference Frame
Projecting onto the Color Image Plane
Visualizing a Point Cloud
Visualizing with PCL
Visualizing with OpenGL
Summary
Chapter 8: Point Clouds, Part 2
Registration
2-D Registration
3-D Registration
Robustness to Outliers
Simultaneous Localization and Mapping (SLAM)
SLAM Using a Conventional Camera
Advantages of Using the Kinect for SLAM
A SLAM Algorithm Using the Kinect
Real-Time Considerations
Surface Reconstruction
Normal Estimation
Triangulation of Points
Summary
Chapter 9: Object Modeling and Detection
Acquiring an Object Model Using a Single Kinect Image
Tabletop Object Detector
Fitting a Parametric Model to a Point Cloud
Building a 3-DModel by Extrusion
Acquiring a 3-D Object Model Using Multiple Views
Overview of a Marker-Based Scanner
Building a Support with Markers
Estimating the 3-DCenter of the Markers in the Camera Space
Kinect Pose Estimation from Markers
Cleaning and Cropping the Partial Views
Merging the Point Clouds
Getting a Better Resolution
Detecting Acquired Objects
Detection Using Global Descriptors
Estimating the Pose of a Recognized Model
Summary
Chapter 10: Multiple Kinects
Why Multiple Kinects?
The Kinect Has a Limited Field of View
The Kinect Fills Data from a Single Direction Only
The Kinect Casts Depth Shadows in Occlusions
What Are the Issues with Multiple Kinects?
Hardware Requirements
Interference Between Kinects
Calibration Between Kinects
Interference
Calibration
Summary
Index
← Prev
Back
Next →
← Prev
Back
Next →