Struct opencv::rgbd::FastICPOdometry [−][src]
pub struct FastICPOdometry { /* fields omitted */ }
Expand description
A faster version of ICPOdometry which is used in KinectFusion implementation Partial list of differences:
- Works in parallel
- Written in universal intrinsics
- Filters points by angle
- Interpolates points and normals
- Doesn’t use masks or min/max depth filtering
- Doesn’t use random subsets of points
- Supports only Rt transform type
- Supports only 4-float vectors as input type
Implementations
Constructor.
Parameters
- cameraMatrix: Camera matrix
- maxDistDiff: Correspondences between pixels of two given frames will be filtered out if their depth difference is larger than maxDepthDiff
- angleThreshold: Correspondence will be filtered out if an angle between their normals is bigger than threshold
- sigmaDepth: Depth sigma in meters for bilateral smooth
- sigmaSpatial: Spatial sigma in pixels for bilateral smooth
- kernelSize: Kernel size in pixels for bilateral smooth
- iterCounts: Count of iterations on each pyramid level
C++ default parameters
- max_dist_diff: Odometry::DEFAULT_MAX_DEPTH_DIFF()
- angle_threshold: (float)(30.*CV_PI/180.)
- sigma_depth: 0.04f
- sigma_spatial: 4.5f
- kernel_size: 7
- iter_counts: std::vector
()
C++ default parameters
- max_dist_diff: Odometry::DEFAULT_MAX_DEPTH_DIFF()
- angle_threshold: (float)(30.*CV_PI/180.)
- sigma_depth: 0.04f
- sigma_spatial: 4.5f
- kernel_size: 7
- iter_counts: std::vector
()
Trait Implementations
Stores algorithm parameters in a file storage
simplified API for language bindings Stores algorithm parameters in a file storage Read more
Returns true if the Algorithm is empty (e.g. in the very beginning or after unsuccessful read
Saves the algorithm to a file. In order to make this method work, the derived class must implement Algorithm::write(FileStorage& fs). Read more
Returns the algorithm string identifier. This string is used as top level xml/yml node tag when the object is saved to a file or string. Read more
fn prepare_frame_cache(
&self,
frame: &mut Ptr<OdometryFrame>,
cache_type: i32
) -> Result<Size>
Performs the conversion.
Method to compute a transformation from the source frame to the destination one. Some odometry algorithms do not used some data of frames (eg. ICP does not use images). In such case corresponding arguments can be set as empty Mat. The method returns true if all internal computations were possible (e.g. there were enough correspondences, system of equations has a solution, etc) and resulting transformation satisfies some test if it’s provided by the Odometry inheritor implementation (e.g. thresholds for maximum translation and rotation). Read more
fn compute2(
&self,
src_frame: &mut Ptr<OdometryFrame>,
dst_frame: &mut Ptr<OdometryFrame>,
rt: &mut dyn ToOutputArray,
init_rt: &Mat
) -> Result<bool>
fn compute2(
&self,
src_frame: &mut Ptr<OdometryFrame>,
dst_frame: &mut Ptr<OdometryFrame>,
rt: &mut dyn ToOutputArray,
init_rt: &Mat
) -> Result<bool>
One more method to compute a transformation from the source frame to the destination one. It is designed to save on computing the frame data (image pyramids, normals, etc.). Read more
fn prepare_frame_cache(
&self,
frame: &mut Ptr<OdometryFrame>,
cache_type: i32
) -> Result<Size>
fn prepare_frame_cache(
&self,
frame: &mut Ptr<OdometryFrame>,
cache_type: i32
) -> Result<Size>
Prepare a cache for the frame. The function checks the precomputed/passed data (throws the error if this data does not satisfy) and computes all remaining cache data needed for the frame. Returned size is a resolution of the prepared frame. Read more
See also Read more
See also Read more