Common Types

Defines common types used by several components. More...

Data Structures
struct	VPIPyramidalKeypointF32
	Stores a float32 pyramidal-based keypoint coordinate The coordinate includes the (x, y) position in an image inside a pyramid, and (octave and layer) selects a specific image on that pyramid. More...
struct	VPIKeypointF32
	Stores a float32 keypoint coordinate The coordinate is relative to the top-left corner of an image. More...
struct	VPIKeypointU32
	Stores a U32 keypoint coordinate The coordinate is relative to the top-left corner of an image. More...
struct	VPIHomographyTransform2D
	Stores a generic 2D homography transform. More...
struct	VPIBoundingBox
	Stores a generic 2D bounding box. More...
struct	VPIAxisAlignedBoundingBoxF32
	Stores an axis-aligned 32-bit floating point 2D bounding box. More...
struct	VPIStats
	Stores the statistics of an image. More...
struct	VPIMatches
	Stores the matches between 2 descriptors. More...
struct	VPIRectangleI
	Stores the geometric information of a rectangle. More...

Macros
#define	VPI_TERMINATION_CRITERIA_ITERATIONS   (1u << 0)
	Defines the termination criteria macros. More...

Enumerations
enum	VPIInterpolationType
	Interpolation types supported by several algorithms. More...
enum	VPIBorderExtension
	Image border extension specify how pixel values outside of the image domain should be constructed. More...
enum	VPIOpticalFlowQuality
	Defines the quality of the optical flow algorithm. More...
enum	VPILockMode
	Defines the lock modes used by memory lock functions. More...
enum	VPIEventState
	Defines the states of the event. More...

Detailed Description

Defines common types used by several components.

---

Data Structure Documentation

VPIPyramidalKeypointF32

struct VPIPyramidalKeypointF32

Stores a float32 pyramidal-based keypoint coordinate The coordinate includes the (x, y) position in an image inside a pyramid, and (octave and layer) selects a specific image on that pyramid.

Pyramids are used to approximate the scale space with one or more resolution levels, also known as octaves, where each octave may have one or more smoothing layers. One example is in the ORB algorithm, where its input is a pyramid, representing the scale space of its base image, that normally has several octaves, where each octave has a single layer. Octaves are defined as levels of dyadic-image pyramids, i.e. a VPIPyramid with scale=0.5.

Definition at line 319 of file Types.h.

Collaboration diagram for VPIPyramidalKeypointF32:

Data Fields
float	x	Keypoint's x coordinate.
float	y	Keypoint's y coordinate.
float	octave	Pyramid octave (or pyramid level) where the keypoint is located.
float	layer	Pyramid layer within the octave where the keypoint is located.

VPIKeypointF32

struct VPIKeypointF32

Stores a float32 keypoint coordinate The coordinate is relative to the top-left corner of an image.

Definition at line 333 of file Types.h.

Collaboration diagram for VPIKeypointF32:

Data Fields
float	x	Keypoint's x coordinate.
float	y	Keypoint's y coordinate.

VPIKeypointU32

struct VPIKeypointU32

Stores a U32 keypoint coordinate The coordinate is relative to the top-left corner of an image.

Definition at line 345 of file Types.h.

Collaboration diagram for VPIKeypointU32:

Data Fields
uint32_t	x	Keypoint's x coordinate.
uint32_t	y	Keypoint's y coordinate.

VPIHomographyTransform2D

struct VPIHomographyTransform2D

Stores a generic 2D homography transform.

When only scale and translation transformation is needed, these parameters must be arranged in the matrix as follows:

\[ \begin{bmatrix} s_x & 0 & p_x \\ 0 & s_y & p_y \\ 0 & 0 & 1 \end{bmatrix} \]

Scaling \((s_x,s_y)\) is relative to the center of the patch, position \((p_x,p_y)\) is relative to the top-left of the image.

In the general case, given an homogeneous 2D point \(P(x,y,1)\) and the matrix \(M^{3x3}\), the Euclidean 2D point \(O(x,y)\) is defined as

\begin{align} T &= M \cdot P \\ O &= (T_x/T_z, T_y/T_z) \end{align}

Definition at line 403 of file Types.h.

Collaboration diagram for VPIHomographyTransform2D:

Data Fields
float	mat3[3][3]	3x3 homogeneous matrix that defines the homography.

VPIBoundingBox

struct VPIBoundingBox

Stores a generic 2D bounding box.

Although this structure can store a 2D bounding box transformed by any homography, sometimes it stores an axis-aligned bounding box. To retrieve it, do the following:

float x = xform.mat3[0][2];
float y = xform.mat3[1][2];
float w = width  * xform.mat3[0][0];
float h = height * xform.mat3[1][1];

Definition at line 423 of file Types.h.

Collaboration diagram for VPIBoundingBox:

Data Fields
VPIHomographyTransform2D	xform	Defines the bounding box top left corner and its homography.
float	width	Bounding box width.
float	height	Bounding box height.

VPIAxisAlignedBoundingBoxF32

struct VPIAxisAlignedBoundingBoxF32

Stores an axis-aligned 32-bit floating point 2D bounding box.

Definition at line 435 of file Types.h.

Collaboration diagram for VPIAxisAlignedBoundingBoxF32:

Data Fields
float	left	Defines the bounding box's left corner's coordinate.
float	top	Defines the bounding box's top corner's coordinate.
float	width	Bounding box width.
float	height	Bounding box height.

VPIStats

struct VPIStats

Stores the statistics of an image.

Definition at line 547 of file Types.h.

Collaboration diagram for VPIStats:

Data Fields
float	mean[4]	Per channel mean.
float	covariance[4][4]	Per-channel covariance matrix, the square root of the diagonal is the standard deviation.
int32_t	pixelCount	Total pixel count.
float	sum[4]	Per channel sum.

VPIMatches

struct VPIMatches

Stores the matches between 2 descriptors.

Definition at line 573 of file Types.h.

Collaboration diagram for VPIMatches:

Data Fields
int	refIndex[VPI_MAX_MATCHES_PER_DESCRIPTOR]	Match reference index.
float	distance[VPI_MAX_MATCHES_PER_DESCRIPTOR]	Distance between query and reference descriptor.

VPIRectangleI

struct VPIRectangleI

Stores the geometric information of a rectangle.

Definition at line 587 of file Types.h.

Collaboration diagram for VPIRectangleI:

Data Fields
int32_t	x
int32_t	y
int32_t	width
int32_t	height

Macro Definition Documentation

VPI_TERMINATION_CRITERIA_ITERATIONS

#define VPI_TERMINATION_CRITERIA_ITERATIONS   (1u << 0)

#include <vpi/Types.h>

Defines the termination criteria macros.

Termination based on maximum number of iterations.

Definition at line 638 of file Types.h.

Enumeration Type Documentation

VPIInterpolationType

enum VPIInterpolationType

#include <vpi/Interpolation.h>

Interpolation types supported by several algorithms.

Enumerator
VPI_INTERP_NEAREST	Nearest neighbor interpolation. \[ P(x,y) = \mathit{in}[\lfloor x+0.5 \rfloor, \lfloor y+0.5 \rfloor] \]
VPI_INTERP_LINEAR	Linear interpolation. Interpolation weights are defined as: \begin{align*} w_0(t)& \triangleq t-\lfloor t \rfloor \\ w_1(t)& \triangleq 1 - w_0(t) \\ \end{align*} Bilinear-interpolated value is given by the formula below: \[ P(x,y) = \sum_{p=0}^1 \sum_{q=0}^1 \mathit{in}[\lfloor x \rfloor+p, \lfloor y \rfloor+q]w_p(x)w_q(y) \]
VPI_INTERP_CATMULL_ROM	Catmull-Rom cubic interpolation. Catmull-Rom interpolation weights with \(A=-0.5\) are defined as follows: \begin{eqnarray*} w_0(t) &\triangleq& A(t+1)^3 &-& 5A(t+1)^2 &+& 8A(t+1) &-& 4A \\ w_1(t) &\triangleq& (A+2)t^3 &-& (A+3)t^2 &\nonumber& &+& 1 \\ w_2(t) &\triangleq& (A+2)(1-t)^3 &-& (A+3)(1-t)^2 &\nonumber& &+& 1 \\ w_3(t) &\triangleq& \rlap{1 - w_0(t) - w_1(t) - w_2(t) } \end{eqnarray*} Bicubic-interpolated value is given by the formula below: \[ P(x,y) = \sum_{p=-1}^2 \sum_{q=-1}^2 \mathit{in}[\lfloor x \rfloor+p, \lfloor y \rfloor+q]w_p(x)w_q(y) \]

Definition at line 69 of file Interpolation.h.

VPIBorderExtension

enum VPIBorderExtension

#include <vpi/Types.h>

Image border extension specify how pixel values outside of the image domain should be constructed.

Enumerator
VPI_BORDER_ZERO	All pixels outside the image are considered to be zero.
VPI_BORDER_CLAMP	Border pixels are repeated indefinitely.
VPI_BORDER_REFLECT	edcba|abcde|edcba
VPI_BORDER_MIRROR	dedcb|abcde|dcbab
VPI_BORDER_LIMITED	Consider image as limited to not access outside pixels.
VPI_BORDER_INVALID	Invalid border.

Definition at line 276 of file Types.h.

VPIOpticalFlowQuality

enum VPIOpticalFlowQuality

#include <vpi/Types.h>

Defines the quality of the optical flow algorithm.

Enumerator
VPI_OPTICAL_FLOW_QUALITY_LOW	Fast but low quality optical flow implementation.
VPI_OPTICAL_FLOW_QUALITY_MEDIUM	Speed and quality in between of VPI_OPTICAL_FLOW_QUALITY_LOW and VPI_OPTICAL_FLOW_QUALITY_HIGH.
VPI_OPTICAL_FLOW_QUALITY_HIGH	Slow but high quality optical flow implementation.

Definition at line 597 of file Types.h.

VPILockMode

enum VPILockMode

#include <vpi/Types.h>

Defines the lock modes used by memory lock functions.

Enumerator
VPI_LOCK_READ	Lock memory only for reading. Writing to the memory when locking for reading leads to undefined behavior.
VPI_LOCK_WRITE	Lock memory only for writing. Reading to the memory when locking for reading leads to undefined behavior. It is expected that the whole memory is written to. If there are regions not written, it might not be updated correctly during unlock. In this case, it's better to use VPI_LOCK_READ_WRITE. It might be slightly efficient to lock only for writing, specially when performing non-shared memory mapping.
VPI_LOCK_READ_WRITE	Lock memory for reading and writing.

Definition at line 613 of file Types.h.

VPIEventState

enum VPIEventState

#include <vpi/Types.h>

Defines the states of the event.

Enumerator
VPI_EVENT_STATE_NOT_SIGNALED	Event is not signaled yet.
VPI_EVENT_STATE_SIGNALED	Event has been signaled.

Definition at line 695 of file Types.h.