Functions for estimation of difference between current frame and background. More...

Functions
SIMD_API void	SimdAddFeatureDifference (const uint8_t value, size_t valueStride, size_t width, size_t height, const uint8_t lo, size_t loStride, const uint8_t hi, size_t hiStride, uint16_t weight, uint8_t difference, size_t differenceStride)
	Adds feature difference to common difference sum. More...

template<template< class > class A>
SIMD_INLINE void	AddFeatureDifference (const View< A > &value, const View< A > &lo, const View< A > &hi, uint16_t weight, View< A > &difference)
	Adds feature difference to common difference sum. More...

Detailed Description

Functions for estimation of difference between current frame and background.

Function Documentation

void SimdAddFeatureDifference	(	const uint8_t *	value,
		size_t	valueStride,
		size_t	width,
		size_t	height,
		const uint8_t *	lo,
		size_t	loStride,
		const uint8_t *	hi,
		size_t	hiStride,
		uint16_t	weight,
		uint8_t *	difference,
		size_t	differenceStride
	)

Adds feature difference to common difference sum.

All images must have the same width, height and format (8-bit gray).

For every point:

excess = max(lo[i] - value[i], 0) + max(value[i] - hi[i], 0);
difference[i] += (weight * excess*excess) >> 16;

This function is used for difference estimation in algorithm of motion detection.

Note: This function has a C++ wrapper Simd::AddFeatureDifference(const View<A>& value, const View<A>& lo, const View<A>& hi, uint16_t weight, View<A>& difference).

Parameters

[in]	value	- a pointer to pixels data of current feature value.
[in]	valueStride	- a row size of the value image.
[in]	width	- an image width.
[in]	height	- an image height.
[in]	lo	- a pointer to pixels data of feature lower bound of dynamic background.
[in]	loStride	- a row size of the lo image.
[in]	hi	- a pointer to pixels data of feature upper bound of dynamic background.
[in]	hiStride	- a row size of the hi image.
[in]	weight	- a current feature weight (unsigned 16-bit value).
[in,out]	difference	- a pointer to pixels data of image with total difference.
[in]	differenceStride	- a row size of difference image.

Adds feature difference to common difference sum.

All images must have the same width, height and format (8-bit gray).

For every point:

excess = max(lo[i] - value[i], 0) + max(value[i] - hi[i], 0);
difference[i] += (weight * excess*excess) >> 16;

This function is used for difference estimation in algorithm of motion detection.

Parameters

[in]	value	- a current feature value.
[in]	lo	- a feature lower bound of dynamic background.
[in]	hi	- a feature upper bound of dynamic background.
[in]	weight	- a current feature weight (unsigned 16-bit value).
[in,out]	difference	- an image with total difference.