Functions related with homography transformation. More...

#include "definitions/define.h"
#include "headers/homography.h"

Include dependency graph for homography.cpp:

Functions
bool	findHomographyMatrix (const cv::Mat &image_src, const cv::Mat &image_dst, cv::Mat &homography_matrix, cv::Mat &ransac_mask)
	Function that find the homography matrix that leaves the imageSrc to the plan of the imageDst. More...

bool	findHomographyMatrix (const cv::Mat &image_src, const cv::Mat &image_dst, cv::Mat &homography_matrix)
	Function that find the homography matrix that leaves the imageSrc to the plan of the imageDst. More...

bool	findHomographyMatrix (const std::vector< cv::KeyPoint > &keypoints_image_src, const std::vector< cv::KeyPoint > &keypoints_image_dst, const cv::Mat &descriptors_image_src, const cv::Mat &descriptors_image_dst, cv::Mat &homography_matrix, cv::Mat &ransac_mask)
	Function that finds the homography matrix given the keypoints and descriptors. More...

bool	findHomographyMatrix (const std::vector< cv::KeyPoint > &keypoints_image_src, const std::vector< cv::KeyPoint > &keypoints_image_dst, const cv::Mat &descriptors_image_src, const cv::Mat &descriptors_image_dst, cv::Mat &homography_matrix)
	Function that find the homography matrix that leaves the imageSrc to the plan of the imageDst. More...

bool	applyHomographyMatrix (const cv::Mat &image_src, const cv::Mat &homography_matrix, cv::Mat &image_result)
	Function that apply homography matrix in a given image. More...

bool	checkHomographyConsistency (const std::vector< cv::Point2f > img_corners, const cv::Mat &homography_matrix)
	Function that checks if after made the homography transformation the corner consistency is maintained. More...

bool	checkHomographyConsistency (const std::vector< cv::Point2f > new_img_corners)
	Function that checks if after made the homography transformation the corner consistency is maintained. More...

double	getAreaRatio (const cv::Mat &image_src, const cv::Mat &homography_matrix, const cv::Rect &frame_limits)
	Function that calculates the loss(%) of the homography transformation in a ROI. More...

HomogCoverage	getHomogCoverage (const cv::Mat &image_src, const cv::Mat &homography_matrix, const cv::Rect &drop_area, const cv::Rect &crop_area)
	Function that calculates the loss(%) of the homography transformation in a ROI. More...

void	getKeypointsAndDescriptors (const cv::Mat &image, std::vector< cv::KeyPoint > &keypoints, cv::Mat &descriptors)
	Function that gets the keypoints and describe an image in order to avoid unnecessary computations for tasks like finding the homography matrix. More...

Detailed Description

Functions related with homography transformation.

Functions find homography matrix between two images. Apply homography matrix. Check if the image corner consistency is held after application of the homography matrix. Check the area ratio between the image after the homography matrix application and the frame boundaries.

Function Documentation

bool applyHomographyMatrix	(	const cv::Mat &	image_src,
		const cv::Mat &	homography_matrix,
		cv::Mat &	image_result
	)

Function that apply homography matrix in a given image.

Parameters

image_src	- image where the homography matrix will be applied.
homography_matrix	- homography matrix.
image_result	- object to save the image after the application of the homography matrix.

Returns: bool true - if the corner consistency is maintained after application of the homography matrix.
bool false - if the consistency is not maintained after application of the homography matrix. In this case the imageResult is a simple copy of the imageSrc.

Author: Michel Melo da Silva

Date: 25/03/2016

bool checkHomographyConsistency	(	const std::vector< cv::Point2f >	img_corners,
		const cv::Mat &	homography_matrix
	)

Function that checks if after made the homography transformation the corner consistency is maintained.

Parameters

img_corners	- image corners in the following order: `img_corners`[0] = 0 , 0 `img_corners`[1] = cols , 0 `img_corners`[2] = 0 , rows `img_corners`[3] = cols , rows
homography_matrix	- homography matrix that will be checked the consistency.

Returns: bool true - if the consistency is maintained.
bool false - if the consistency is not maintained.

Author: Michel Melo da Silva

Date: 08/04/2016

bool checkHomographyConsistency ( const std::vector< cv::Point2f > new_img_corners )

Function that checks if after made the homography transformation the corner consistency is maintained.

Parameters

new_img_corners - image corners in the following order:
img_corners[0] = 0 , 0
img_corners[1] = cols , 0
img_corners[2] = 0 , rows
img_corners[3] = cols , rows

Returns: bool true - if the consistency is maintained.
bool false - if the consistency is not maintained.

Author: Michel Melo da Silva

Date: 12/04/2016

bool findHomographyMatrix	(	const cv::Mat &	image_src,
		const cv::Mat &	image_dst,
		cv::Mat &	homography_matrix,
		cv::Mat &	ransac_mask
	)

Function that find the homography matrix that leaves the imageSrc to the plan of the imageDst.

Function that find the homography matrix that leave the imageSrc to the plan of the imageDst.

Returns the Ransak mask after finding the homography matrix.

Parameters

image_src	- image to where the homography will be calculated.
image_dst	- image of the destination of the homography.
homography_matrix	- object to homography matrix calculated.
ransac_mask	- RANSAC mask of the homography matrix. 1 means inliers and 0 means outliers.

Returns: bool true - if there is enough points in both images and good matches between them to find a homography matrix.
bool false - if there is not enough points in both images or good matches between them to find a homography matrix.

Author: Michel Melo da Silva

Date: 28/03/2016

bool findHomographyMatrix	(	const cv::Mat &	image_src,
		const cv::Mat &	image_dst,
		cv::Mat &	homography_matrix
	)

Function that find the homography matrix that leaves the imageSrc to the plan of the imageDst.

Function that find the homography matrix that leave the imageSrc to the plan of the imageDst.

Parameters

image_src	- image to where the homography will be calculated.
image_dst	- image of the destination of the homography.
homography_matrix	- object to save the homography matrix calculated.

Returns: bool true - if there is enough points in both images and good matches between them to find a homography matrix.
bool false - if there is not enough points in both images or good matches between them to find a homography matrix.

Author: Michel Melo da Silva

Date: 28/03/2016

bool findHomographyMatrix	(	const std::vector< cv::KeyPoint > &	keypoints_image_src,
		const std::vector< cv::KeyPoint > &	keypoints_image_dst,
		const cv::Mat &	descriptors_image_src,
		const cv::Mat &	descriptors_image_dst,
		cv::Mat &	homography_matrix,
		cv::Mat &	ransac_mask
	)

Function that finds the homography matrix given the keypoints and descriptors.

Returns the Ransak mask after finding the homography matrix.

Parameters

keypoints_image_src	- keypoints of the source image.
keypoints_image_dst	- keypoints of the target image.
descriptors_image_src	- descriptors of the source image.
descriptors_image_dst	- descriptors of the target image.
homography_matrix	- object to homography matrix calculated.
ransac_mask	- RANSAC mask of the homography matrix. 1 means inliers and 0 means outliers.

Returns: bool true - if there is enough points in both images and good matches between them to find a homography matrix.
bool false - if there is not enough points in both images or good matches between them to find a homography matrix.

Author: Washington Luis de Souza Ramos

Date: 30/08/2016

bool findHomographyMatrix	(	const std::vector< cv::KeyPoint > &	keypoints_image_src,
		const std::vector< cv::KeyPoint > &	keypoints_image_dst,
		const cv::Mat &	descriptors_image_src,
		const cv::Mat &	descriptors_image_dst,
		cv::Mat &	homography_matrix
	)

Function that find the homography matrix that leaves the imageSrc to the plan of the imageDst.

Parameters

keypoints_image_src	- keypoints of the source image.
keypoints_image_dst	- keypoints of the target image.
descriptors_image_src	- descriptors of the source image.
descriptors_image_dst	- descriptors of the target image.
homography_matrix	- object to save the homography matrix calculated.

Returns: bool true - if there is enough points in both images and good matches between them to find a homography matrix.
bool false - if there is not enough points in both images or good matches between them to find a homography matrix.

Author: Washington Luis de Souza Ramos

Date: 09/09/2016

double getAreaRatio	(	const cv::Mat &	image_src,
		const cv::Mat &	homography_matrix,
		const cv::Rect &	frame_limits
	)

Function that calculates the loss(%) of the homography transformation in a ROI.

It returns the ratio between the non-image area and the frame_limits area.

Parameters

image_src	- The source image.
homography_matrix	- The homography matrix.
frame_limits	- The region of interest (ROI).

Returns: double - The ratio between non-image area and the whole frame area.

Author: Washington Luis de Souza Ramos

Date: 14/04/2016

HomogCoverage getHomogCoverage	(	const cv::Mat &	image_src,
		const cv::Mat &	homography_matrix,
		const cv::Rect &	drop_area,
		const cv::Rect &	crop_area
	)

Function that calculates the loss(%) of the homography transformation in a ROI.

It returns the coverage of a transformation which is NONE, DROP_AREA or CROP_AREA.

Parameters

image_src	- The source image.
homography_matrix	- The homography matrix.
drop_area	- The region of interest (ROI) for the drop area.
crop_area	- The region of interest (ROI) for the crop area.

Returns: The HomogCoverage enum value

Author: Washington Luis de Souza Ramos

Date: 10/09/2016

void getKeypointsAndDescriptors	(	const cv::Mat &	image,
		std::vector< cv::KeyPoint > &	keypoints,
		cv::Mat &	descriptors
	)

Function that gets the keypoints and describe an image in order to avoid unnecessary computations for tasks like finding the homography matrix.

Parameters

image	- image to where the keypoints and descriptors are to be extracted from.
keypoints	- the keypoints found in the image.
descriptors	- the descriptors of the image.

Author: Washington Luis de Souza Ramos

Date: 30/08/2016