Template Matching Algorithm

Runs the template matching algorithm with template image over the searched image. More...

Enumerations
enum	VPITemplateMatchingMethod
	Define method to calculate the template matching score. More...

Functions
VPIStatus	vpiCreateTemplateMatching (uint64_t backends, int32_t imageWidth, int32_t imageHeight, VPIPayload *payload)
	Creates payload for vpiSubmitTemplateMatching. More...
VPIStatus	vpiTemplateMatchingSetSourceImage (VPIStream stream, uint64_t backend, VPIPayload payload, VPIImage srcImage)
	Set the source image. More...
VPIStatus	vpiTemplateMatchingSetTemplateImage (VPIStream stream, uint64_t backend, VPIPayload payload, VPIImage templImage, VPIImage mask)
	Set the template image. More...
VPIStatus	vpiSubmitTemplateMatching (VPIStream stream, uint64_t backend, VPIPayload payload, VPIImage output, VPITemplateMatchingMethod method)
	Runs the template matching algorithm with provided template. More...

Detailed Description

Runs the template matching algorithm with template image over the searched image.

Refer to Template Matching for more details and usage examples.

Enumeration Type Documentation

VPITemplateMatchingMethod

enum VPITemplateMatchingMethod

#include </opt/nvidia/vpi3/include/vpi/algo/TemplateMatching.h>

Define method to calculate the template matching score.

Enumerator
VPI_TEMPLATE_MATCHING_NCC	Normalized cross correlation.

Definition at line 195 of file TemplateMatching.h.

Function Documentation

vpiCreateTemplateMatching()

VPIStatus vpiCreateTemplateMatching	(	uint64_t	backends,
		int32_t	imageWidth,
		int32_t	imageHeight,
		VPIPayload *	payload
	)

#include </opt/nvidia/vpi3/include/vpi/algo/TemplateMatching.h>

Creates payload for vpiSubmitTemplateMatching.

Attention

For this function to succeed, the below libraries must to be installed in the system:

• libnppc.so.11
• libnppial.so.11
• libnppidei.so.11
• libnppist.so.11

Parameters

[in]	backends	VPI backends that are eligible to execute the algorithm. Valid values: VPI_BACKEND_CPU VPI_BACKEND_CUDA Backend must be enabled in current context.
[in]	imageWidth	Source image width.
[in]	imageHeight	Source image height.
[out]	payload	Pointer to the payload variable that receives the created handle.

Return values

VPI_ERROR_INVALID_ARGUMENT	payload is NULL.
VPI_ERROR_INVALID_ARGUMENT	imageWidth and imageHeight outside valid range.
VPI_ERROR_INVALID_ARGUMENT	backends refers to an invalid backend.
VPI_ERROR_INVALID_OPERATION	Backend isn't enabled in current context.
VPI_ERROR_INVALID_OPERATION	Required NPP libraries aren't installed in the system.
VPI_ERROR_NOT_IMPLEMENTED	Algorithm is not supported by given backend.
VPI_ERROR_INVALID_CONTEXT	Current context is destroyed.
VPI_ERROR_OUT_OF_MEMORY	Cannot allocate required resources.
VPI_SUCCESS	Operation executed successfully.

vpiTemplateMatchingSetSourceImage()

VPIStatus vpiTemplateMatchingSetSourceImage	(	VPIStream	stream,
		uint64_t	backend,
		VPIPayload	payload,
		VPIImage	srcImage
	)

#include </opt/nvidia/vpi3/include/vpi/algo/TemplateMatching.h>

Set the source image.

Parameters

[in]	stream	A stream handle where the operation will be queued into. Must not be NULL. Stream must have enabled the backends that will execute the algorithm.
[in]	backend	Backend that will execute the algorithm. Valid values: VPI_BACKEND_CPU VPI_BACKEND_CUDA
[in]	payload	Pointer to the payload variable that receives the created handle.
[in]	srcImage	Source image where the template image will be searching on. Must not be NULL. Image must have enabled the backends that will execute the algorithm. The accepted image formats are: Format CPU CUDA VPI_IMAGE_FORMAT_U8 * * VPI_IMAGE_FORMAT_Y8_ER * * VPI_IMAGE_FORMAT_Y8 * *

Return values

VPI_ERROR_INVALID_ARGUMENT	payload is NULL.
VPI_ERROR_INVALID_ARGUMENT	srcImage is NULL.
VPI_ERROR_INVALID_ARGUMENT	srcImage resolution does not match with the one specified in the vpiCreateTemplateMatching.
VPI_ERROR_INVALID_OPERATION	Backend isn't enabled in current context.
VPI_ERROR_INVALID_OPERATION	The needed backends aren't enabled in srcImage.
VPI_ERROR_INVALID_CONTEXT	Current context is destroyed.
VPI_ERROR_OUT_OF_MEMORY	Cannot allocate required resources.
VPI_SUCCESS	Operation executed successfully.

vpiTemplateMatchingSetTemplateImage()

VPIStatus vpiTemplateMatchingSetTemplateImage	(	VPIStream	stream,
		uint64_t	backend,
		VPIPayload	payload,
		VPIImage	templImage,
		VPIImage	mask
	)

#include </opt/nvidia/vpi3/include/vpi/algo/TemplateMatching.h>

Set the template image.

Parameters

[in]	stream	A stream handle where the operation will be queued into. Must not be NULL. Stream must have enabled the backends that will execute the algorithm.
[in]	backend	Backend that will execute the algorithm. Valid values: VPI_BACKEND_CPU VPI_BACKEND_CUDA
[in]	payload	Pointer to the payload variable that receives the created handle.
[in]	templImage	Template image. Must not be NULL. Image must have enabled the backends that will execute the algorithm. The accepted image formats are: Format CPU CUDA VPI_IMAGE_FORMAT_U8 * * VPI_IMAGE_FORMAT_Y8_ER * * VPI_IMAGE_FORMAT_Y8 * *
[in]	mask	Mask used when calculating the template matching score. Could be NULL if not needed. Must have same dimension as the template image The accepted image formats are: Format CPU CUDA VPI_IMAGE_FORMAT_U8 * *

Return values

VPI_ERROR_INVALID_ARGUMENT	payload is NULL.
VPI_ERROR_INVALID_ARGUMENT	templImage is NULL.
VPI_ERROR_INVALID_ARGUMENT	templImage resolution is greater than source image.
VPI_ERROR_INVALID_OPERATION	Backend isn't enabled in current context.
VPI_ERROR_INVALID_OPERATION	The needed backends aren't enabled in templImage.
VPI_ERROR_INVALID_CONTEXT	Current context is destroyed.
VPI_ERROR_OUT_OF_MEMORY	Cannot allocate required resources.
VPI_SUCCESS	Operation executed successfully.

vpiSubmitTemplateMatching()

VPIStatus vpiSubmitTemplateMatching	(	VPIStream	stream,
		uint64_t	backend,
		VPIPayload	payload,
		VPIImage	output,
		VPITemplateMatchingMethod	method
	)

#include </opt/nvidia/vpi3/include/vpi/algo/TemplateMatching.h>

Runs the template matching algorithm with provided template.

Parameters

[in]	stream	A stream handle where the operation will be queued into. Must not be NULL. Stream must have enabled the backends that will execute the algorithm.
[in]	backend	Backend that will execute the algorithm. Valid values: VPI_BACKEND_CPU VPI_BACKEND_CUDA
[in]	payload	Pointer to the payload variable that receives the created handle.
[out]	output	Output image where the template matching score is written to. Must not be NULL. Assume the source image is with resolution WxH and template image is with resolution wxh, then the output image shall have resolution (W - w + 1) * (H - h + 1) Image format has to be VPI_IMAGE_FORMAT_F32. Image must have enabled the backends that will execute the algorithm.
[in]	method	Specify the method when calculating the template matching score.

Return values

VPI_ERROR_INVALID_ARGUMENT	stream is NULL.
VPI_ERROR_INVALID_ARGUMENT	output is NULL.
VPI_ERROR_INVALID_ARGUMENT	backend is invalid or unknown.
VPI_ERROR_INVALID_IMAGE_FORMAT	output image format not supported.
VPI_ERROR_INVALID_OPERATION	Backend hardware is not available.
VPI_ERROR_INVALID_OPERATION	The needed backends aren't enabled in stream and output.
VPI_ERROR_NOT_IMPLEMENTED	Algorithm is not supported by given backend.
VPI_SUCCESS	Operation executed successfully.