Compute Library
NEHarrisScoreKernel< block_size > Class Template Reference

Template NEON kernel to perform Harris Score. More...

#include <NEHarrisCornersKernel.h>

Collaboration diagram for NEHarrisScoreKernel< block_size >:

Public Member Functions

const char * name () const override
 Name of the kernel. More...
 NEHarrisScoreKernel ()
 Default constructor. More...
void configure (const IImage *input1, const IImage *input2, IImage *output, float norm_factor, float strength_thresh, float sensitivity, bool border_undefined) override
 Setup the kernel parameters. More...
BorderSize border_size () const override
 The size of the border for that kernel. More...
void run (const Window &window, const ThreadInfo &info) override
 Execute the kernel on the passed window. More...
- Public Member Functions inherited from INEHarrisScoreKernel
 INEHarrisScoreKernel ()
 Default constructor. More...
 INEHarrisScoreKernel (const INEHarrisScoreKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
INEHarrisScoreKerneloperator= (const INEHarrisScoreKernel &)=delete
 Prevent instances of this class from being copied (As this class contains pointers) More...
 INEHarrisScoreKernel (INEHarrisScoreKernel &&)=default
 Allow instances of this class to be moved. More...
INEHarrisScoreKerneloperator= (INEHarrisScoreKernel &&)=default
 Allow instances of this class to be moved. More...
 ~INEHarrisScoreKernel ()=default
 Default destructor. More...
- Public Member Functions inherited from ICPPKernel
virtual ~ICPPKernel ()=default
 Default destructor. More...
- Public Member Functions inherited from IKernel
 IKernel ()
 Constructor. More...
virtual ~IKernel ()=default
 Destructor. More...
virtual bool is_parallelisable () const
 Indicates whether or not the kernel is parallelisable. More...
const Windowwindow () const
 The maximum window the kernel can be executed on. More...

Detailed Description

template<int32_t block_size>
class arm_compute::NEHarrisScoreKernel< block_size >

Template NEON kernel to perform Harris Score.

The implementation supports 3, 5, and 7 for the block_size

Definition at line 84 of file NEHarrisCornersKernel.h.

Constructor & Destructor Documentation

◆ NEHarrisScoreKernel()

Default constructor.

Definition at line 703 of file NEHarrisCornersKernel.cpp.

704  : INEHarrisScoreKernel(), _func(nullptr)
705 {
706 }

Member Function Documentation

◆ border_size()

BorderSize border_size ( ) const

The size of the border for that kernel.

The width in number of elements of the border.

Reimplemented from IKernel.

Definition at line 730 of file NEHarrisCornersKernel.cpp.

731 {
732  return _border_size;
733 }

◆ configure()

void configure ( const IImage input1,
const IImage input2,
IImage output,
float  norm_factor,
float  strength_thresh,
float  sensitivity,
bool  border_undefined 

Setup the kernel parameters.

[in]input1Source image (gradient X). Data types supported: S16/S32
[in]input2Source image (gradient Y). Data types supported: same as @ input1
[out]outputDestination image (harris score). Data types supported: F32
[in]norm_factorNormalization factor to use accordingly with the gradient size (Must be different from 0)
[in]strength_threshMinimum threshold with which to eliminate Harris Corner scores (computed using the normalized Sobel kernel).
[in]sensitivitySensitivity threshold k from the Harris-Stephens equation
[in]border_undefinedTrue if the border mode is undefined. False if it's replicate or constant.

Implements INEHarrisScoreKernel.

Definition at line 736 of file NEHarrisCornersKernel.cpp.

738 {
746  ARM_COMPUTE_ERROR_ON(0.0f == norm_factor);
748  _input1 = input1;
749  _input2 = input2;
750  _output = output;
751  _sensitivity = sensitivity;
752  _strength_thresh = strength_thresh;
753  _norm_factor = norm_factor;
754  _border_size = BorderSize(block_size / 2);
756  if(input1->info()->data_type() == DataType::S16)
757  {
758  switch(block_size)
759  {
760  case 3:
761  _func = &harris_score3x3_S16_S16_FLOAT;
762  break;
763  case 5:
764  _func = &harris_score5x5_S16_S16_FLOAT;
765  break;
766  case 7:
767  _func = &harris_score7x7_S16_S16_FLOAT;
768  break;
769  default:
770  ARM_COMPUTE_ERROR("Invalid block size");
771  break;
772  }
773  }
774  else
775  {
776  switch(block_size)
777  {
778  case 3:
779  _func = &harris_score3x3_S32_S32_FLOAT;
780  break;
781  case 5:
782  _func = &harris_score5x5_S32_S32_FLOAT;
783  break;
784  case 7:
785  _func = &harris_score7x7_S32_S32_FLOAT;
786  break;
787  default:
788  ARM_COMPUTE_ERROR("Invalid block size");
789  break;
790  }
791  }
793  ARM_COMPUTE_ERROR_ON(nullptr == _func);
795  constexpr unsigned int num_elems_processed_per_iteration = block_size != 7 ? 8 : 4;
796  constexpr unsigned int num_elems_read_per_iteration = block_size != 7 ? 16 : 12;
797  constexpr unsigned int num_elems_written_per_iteration = block_size != 7 ? 8 : 4;
798  constexpr unsigned int num_rows_read_per_iteration = block_size;
800  // Configure kernel window
801  Window win = calculate_max_window(*input1->info(), Steps(num_elems_processed_per_iteration), border_undefined, border_size());
802  AccessWindowHorizontal output_access(output->info(), 0, num_elems_written_per_iteration);
805  AccessWindowRectangle(input1->info(), -_border_size.left,, num_elems_read_per_iteration, num_rows_read_per_iteration),
806  AccessWindowRectangle(input2->info(), -_border_size.left,, num_elems_read_per_iteration, num_rows_read_per_iteration),
807  output_access);
810  input2->info()->valid_region());
812  output_access.set_valid_region(win, valid_region, border_undefined, border_size());
814  INEKernel::configure(win);
815 }
Definition: Validate.h:543
Container for 2D border size.
Definition: Types.h:268
#define ARM_COMPUTE_ERROR(msg)
Print the given message then throw an std::runtime_error.
Definition: Error.h:352
virtual DataType data_type() const =0
Data type used for each element of the tensor.
1 channel, 1 F32 per channel
#define ARM_COMPUTE_ERROR_ON(cond)
If the condition is true then an error message is printed and an exception thrown.
Definition: Error.h:466
Window calculate_max_window(const ValidRegion &valid_region, const Steps &steps=Steps(), bool skip_border=false, BorderSize border_size=BorderSize())
Calculate the maximum window for a given tensor shape and border setting.
Definition: Helpers.cpp:28
virtual ValidRegion valid_region() const =0
Valid region of the tensor.
1 channel, 1 S32 per channel
Definition: Validate.h:856
Implementation of a rectangular access pattern.
bool update_window_and_padding(Window &win, Ts &&... patterns)
Update window and padding size for each of the access patterns.
Definition: Helpers.h:402
Class to describe a number of elements in each dimension.
Definition: Steps.h:40
BorderSize border_size() const override
The size of the border for that kernel.
Implementation of a row access pattern.
virtual ITensorInfo * info() const =0
Interface to be implemented by the child class to return the tensor's metadata.
ValidRegion intersect_valid_regions(const Ts &... regions)
Intersect multiple valid regions.
Definition: Helpers.h:503
1 channel, 1 S16 per channel
Definition: Validate.h:790
Container for valid region of a window.
Definition: Types.h:183
Describe a multidimensional execution window.
Definition: Window.h:39


◆ name()

const char* name ( ) const

Name of the kernel.

Kernel name

Implements ICPPKernel.

Definition at line 87 of file NEHarrisCornersKernel.h.

88  {
89  return "NEHarrisScoreKernel";
90  }

◆ run()

void run ( const Window window,
const ThreadInfo info 

Execute the kernel on the passed window.

If is_parallelisable() returns false then the passed window must be equal to window()
The window has to be a region within the window returned by the window() method
The width of the window has to be a multiple of num_elems_processed_per_iteration().
[in]windowRegion on which to execute the kernel. (Must be a region of the window returned by window())
[in]infoInfo about executing thread and CPU.

Implements ICPPKernel.

Definition at line 709 of file NEHarrisCornersKernel.cpp.

710 {
714  ARM_COMPUTE_ERROR_ON(_func == nullptr);
716  Iterator input1(_input1, window);
717  Iterator input2(_input2, window);
718  Iterator output(_output, window);
720  const size_t input_stride = _input1->info()->strides_in_bytes()[1] / element_size_from_data_type(_input1->info()->data_type());
723  {
724  (*_func)(input1.ptr(), input2.ptr(), output.ptr(), input_stride, _norm_factor, _sensitivity, _strength_thresh);
725  },
726  input1, input2, output);
727 }
const Window & window() const
The maximum window the kernel can be executed on.
Definition: IKernel.cpp:28
size_t element_size_from_data_type(DataType dt)
The size in bytes of the data type.
Definition: Utils.h:186
#define ARM_COMPUTE_ERROR_ON(cond)
If the condition is true then an error message is printed and an exception thrown.
Definition: Error.h:466
To avoid unused variables warnings.
Definition: Error.h:152
Coordinates of an item.
Definition: Coordinates.h:37
void execute_window_loop(const Window &w, L &&lambda_function, Ts &&... iterators)
Iterate through the passed window, automatically adjusting the iterators and calling the lambda_funct...
Definition: Helpers.inl:123
Iterator updated by execute_window_loop for each window element.
Definition: Helpers.h:318
Definition: Validate.h:205
Definition: Validate.h:941


The documentation for this class was generated from the following files: