CCL#

Functions#

NVCVStatus pvaCCLCreate(NVCVOperatorHandle *handle, const NVCVTensorRequirements *imageParams, const NVCVTensorRequirements *labelParams, const PVAConnectivityType connectivity, const uint8_t threshold)

Constructs an instance of the CCL operator.

NVCVStatus pvaCCLSubmit(NVCVOperatorHandle handle, cudaStream_t stream, NVCVTensorHandle image, NVCVTensorHandle labels, NVCVTensorHandle numLabels)

Submits the CCL operator to a CUDA stream.

NVCVStatus pvaCCLSubmit(NVCVOperatorHandle handle, cupvaStream_t stream, NVCVTensorHandle image, NVCVTensorHandle labels, NVCVTensorHandle numLabels)

Submits the CCL operator to a cuPVA stream.

Functions#

NVCVStatus pvaCCLCreate(
NVCVOperatorHandle *handle,
const NVCVTensorRequirements *imageParams,
const NVCVTensorRequirements *labelParams,
const PVAConnectivityType connectivity,
const uint8_t threshold,
)#

Constructs an instance of the CCL operator.

Limitations:

Input imageParams: Data Layout: [CHW], [NCHW] only when C = 1 and N = 1

Data Type

Allowed

8bit Unsigned

Yes

8bit Signed

No

16bit Unsigned

No

16bit Signed

No

32bit Unsigned

No

32bit Signed

No

32bit Float

No

64bit Float

No

Output labelParams: Data Layout: [CHW], [NCHW] only when C = 1 and N = 1

Data Type

Allowed

8bit Unsigned

No

8bit Signed

No

16bit Unsigned

Yes

16bit Signed

No

32bit Unsigned

No

32bit Signed

No

32bit Float

No

64bit Float

No

Input/Output Dependency:

Property

Input == Output

Data Layout

Yes

Data Type

No

Number

Yes

Channels

Yes

Width

Yes

Height

Yes

Value 0 in the output labeled image is background. Maximum number of labels: 65535 excluding label 0 which is reserved for background.

Parameters:

  • handle – [out] Where the operator instance handle will be written to.

    • Must not be NULL.

  • imageParams – [in] Pointer to the NVCVTensorRequirements structure which contains input tensor rank, shape, layout and data type information.

  • labelParams – [in] Pointer to the NVCVTensorRequirements structure which contains output tensor rank, shape, layout and data type information.

  • connectivity – [in] Must be PVA_4_CONNECTIVITY or PVA_8_CONNECTIVITY. PVA_4_CONNECTIVITY for 4-connectivity and PVA_8_CONNECTIVITY for 8-connectivity.

  • threshold – [in] Only pixels with values greater than this threshold are treated as foreground when running connected components. If input image is already binary, set it to 0.

Return values:

  • NVCV_ERROR_INVALID_ARGUMENT – Handle is null.

  • NVCV_ERROR_OUT_OF_MEMORY – Not enough memory to create the operator.

  • NVCV_SUCCESS – Operation executed successfully.

NVCVStatus pvaCCLSubmit(
NVCVOperatorHandle handle,
cudaStream_t stream,
NVCVTensorHandle image,
NVCVTensorHandle labels,
NVCVTensorHandle numLabels,
)#

Submits the CCL operator to a CUDA stream.

Parameters:

  • handle – [in] Handle to the operator.

    • Must not be NULL.

  • stream – [in] Handle to a valid CUDA stream.

  • image – [in] Input tensor handle of input image.

  • labels – [out] Output tensor handle of labeled image. Note that labels in the output are not required to be sequential.

  • numLabels – [out] Output tensor handle of number of labels excluding label 0 which is reserved for background. Data Layout: [W] Width: 1 Data Type: NVCV_DATA_TYPE_S32 If the runtime labels exceed the limit (65536 labels including label 0 which is reserved for background), numLabels will return -1.

Return values:

  • NVCV_ERROR_INVALID_ARGUMENT – Some parameter is outside the valid range.

  • NVCV_ERROR_INTERNAL – Internal error in the operator, invalid types passed in.

  • NVCV_SUCCESS – Operation executed successfully.

NVCVStatus pvaCCLSubmit(
NVCVOperatorHandle handle,
cupvaStream_t stream,
NVCVTensorHandle image,
NVCVTensorHandle labels,
NVCVTensorHandle numLabels,
)#

Submits the CCL operator to a cuPVA stream.

Parameters:

  • handle – [in] Handle to the operator.

    • Must not be NULL.

  • stream – [in] Handle to a valid cuPVA stream.

  • image – [in] Input tensor handle of input image.

  • labels – [out] Output tensor handle of labeled image. Note that labels in the output are not required to be sequential

  • numLabels – [out] Output tensor handle of number of labels excluding label 0 which is reserved for background. Data Layout: [W] Width: 1 Data Type: NVCV_DATA_TYPE_S32 If the runtime labels exceed the limit (65536 labels including label 0 which is reserved for background), numLabels will return -1.

Return values:

  • NVCV_ERROR_INVALID_ARGUMENT – Some parameter is outside the valid range.

  • NVCV_ERROR_INTERNAL – Internal error in the operator, invalid types passed in.

  • NVCV_SUCCESS – Operation executed successfully.