libfovis
Classes | Typedefs
Fovis core

Core data structures and algorithms. More...

Classes

struct  CameraIntrinsicsParameters
 Intrinsic parameters for a pinhole camera with plumb-bob distortion model. More...
 
class  DepthSource
 Provides depth estimates for input image pixels. More...
 
class  FeatureMatch
 Represents a single image feature matched between two camera images taken at different times. More...
 
class  IntensityDescriptorExtractor
 Extracts mean-normalized intensity patch around a pixel. More...
 
class  KeypointData
 Image feature used for motion estimation. More...
 
class  MotionEstimator
 Does the heavy lifting for frame-to-frame motion estimation. More...
 
class  OdometryFrame
 Stores data specific to an input frame. More...
 
class  PyramidLevel
 One level of a Gaussian image pyramid. More...
 
class  Rectification
 Maps image coordinates from an input image to image coordinates on a rectified camera. More...
 
class  VisualOdometry
 Main visual odometry class. More...
 

Typedefs

typedef std::map< std::string, std::string > VisualOdometryOptions
 Options. More...
 

Detailed Description

Core data structures and algorithms.

Typedef Documentation

VisualOdometryOptions

Options.

VisualOdometryOptions is a key-value dictionary of user-adjustable options that affect the behavior of the visual odometry algorithm. Keys and values are both expressed as strings

Current options are:

*   "feature-window-size"
*     Type:        Integer
*     Default:     9
*     Range:       1+
*     Description: The size of the n x n image patch surrounding each feature, used for
*                  keypoint matching.
*      
*   "max-pyramid-level"
*     Type:        Integer
*     Default:     3
*     Range:       1+
*     Description: The maximum Gaussian pyramid level to process the image at.
*                  Pyramid level 1 corresponds to the original image.
*
*   "inlier-max-reprojection-error"
*     Type:        Double
*     Default:     1.5
*     Range:       0+
*     Description: The maximum image-space reprojection error (in pixels) a
*                  feature match is allowed to have and still be considered an
*                  inlier in the set of features used for motion estimation.
*
*   "clique-inlier-threshold"
*     Type:        Double
*     Default:     0.1
*     Range:       0+
*     Description: See Howard's greedy max-clique algorithm for determining the 
*                  maximum set of mutually consisten feature matches.  This
*                  specifies the compatibility threshold, in meters.
*
*   "min-features-for-estimate"
*     Type:        Integer
*     Default:     10
*     Range:       0+
*     Description: Minimum number of features in the inlier set for the motion estimate
*     to be considered valid.
*
*   "max-mean-reprojection-error"
*     Type:        Double
*     Default:     10.0
*     Range:       0+
*     Description: Maximum mean reprojection error over the inlier feature matches for the
*     motion estimate to be considered valid.
*
*   "use-subpixel-refinement"
*     Type:        Integer
*     Default:     1
*     Range:       0 or 1
*     Description: Specifies whether or not to refine feature matches to
*     subpixel resolution.
*
*   "feature-search-window"
*     Type:        Integer
*     Default:     25
*     Range:       0+
*     Description: Specifies the size of the search window to apply when
*     searching for feature matches across time frames.  The search is conducted
*     around the feature location predicted by the initial rotation estimate.
*
*   "target-pixels-per-feature"
*     Type:        Integer
*     Default:     250
*     Range:       1+
*     Description: Specifies the desired feature density as a ratio of input
*                  image pixels per feature detected.  This number is used to control
*                  the adaptive feature thresholding.
*
*   "update-target-features-with-refined"
*     Type:        Integer
*     Default:     0
*     Range:       0 or 1
*     Description: When subpixel refinement is enabled, the refined feature
*                  locations can be saved over the original feature
*                  locations.  This has a slightly negative impact on
*                  frame-to-frame visual odometry, but is likely better when
*                  using this library as part of a visual SLAM
*                  algorithm.
*