kernels.cu File Reference

Functions
__global__ void	findMotionKernel (int *image1, int *image2, int *result, int threshold)
	Basic motion finding algorithm which does not take into account the motion of the robot.
__global__ void	findClustersKernel (int *a, int *result)
	Search around each and every pixel in the motion image summing the result to give an indication of the relative density of motion around that pixel.
void	findMotionOnDevice (int *image1, int *image2, int *result, int threshold)
	Wrapper for findMotionKernel.
void	findClustersOnDevice (int *image, int *result)
	Wrapper for findClustersKernel.

Function Documentation

__global__ void findClustersKernel	(	int *	a,
		int *	result
	)

Search around each and every pixel in the motion image summing the result to give an indication of the relative density of motion around that pixel.

Parameters

[in]	a
[out]	result

void findClustersOnDevice	(	int *	image,
		int *	result
	)

Wrapper for findClustersKernel.

Parameters

[in]	a
[out]	result

__global__ void findMotionKernel	(	int *	image1,
		int *	image2,
		int *	result,
		int	threshold
	)

Basic motion finding algorithm which does not take into account the motion of the robot.

Note

Consecutive images are subtracted and thresholded to create a black and white image showing which areas in an image have changed significantly.

Parameters

[in]	image1
[in]	image2
[in]	threshold
[out]	result

void findMotionOnDevice	(	int *	image1,
		int *	image2,
		int *	result,
		int	threshold
	)

Wrapper for findMotionKernel.

Parameters

[in]	image1
[in]	image2
[in]	threshold
[out]	result