The module brings implementations of intensity transformation algorithms to adjust image contrast.

Functions
void	cv::intensity_transform::autoscaling (const Mat input, Mat &output)
	Given an input bgr or grayscale image, apply autoscaling on domain [0, 255] to increase the contrast of the input image and return the resulting image. More...
void	cv::intensity_transform::BIMEF (InputArray input, OutputArray output, float mu=0.5f, float a=-0.3293f, float b=1.1258f)
	Given an input color image, enhance low-light images using the BIMEF method ([276] [277]). More...
void	cv::intensity_transform::BIMEF (InputArray input, OutputArray output, float k, float mu, float a, float b)
	Given an input color image, enhance low-light images using the BIMEF method ([276] [277]). More...
void	cv::intensity_transform::contrastStretching (const Mat input, Mat &output, const int r1, const int s1, const int r2, const int s2)
	Given an input bgr or grayscale image, apply linear contrast stretching on domain [0, 255] and return the resulting image. More...
void	cv::intensity_transform::gammaCorrection (const Mat input, Mat &output, const float gamma)
	Given an input bgr or grayscale image and constant gamma, apply power-law transformation, a.k.a. gamma correction to the image on domain [0, 255] and return the resulting image. More...
void	cv::intensity_transform::logTransform (const Mat input, Mat &output)
	Given an input bgr or grayscale image and constant c, apply log transformation to the image on domain [0, 255] and return the resulting image. More...

Detailed Description

Namespace for all functions is cv::intensity_transform.

Supported Algorithms

• Autoscaling
• Log Transformations
• Power-Law (Gamma) Transformations
• Contrast Stretching
• BIMEF, A Bio-Inspired Multi-Exposure Fusion Framework for Low-light Image Enhancement [276] [277]

References from following book and websites:

• Digital Image Processing 4th Edition Chapter 3 [Rafael C. Gonzalez, Richard E. Woods] [193]
• http://www.cs.uregina.ca/Links/class-info/425/Lab3/ [129]
• https://theailearner.com/2019/01/30/contrast-stretching/ [238]

Function Documentation

autoscaling()

void cv::intensity_transform::autoscaling	(	const Mat	input,
		Mat &	output
	)

Python:
	None	=	cv.intensity_transform.autoscaling(	input, output	)

#include <opencv2/intensity_transform.hpp>

Given an input bgr or grayscale image, apply autoscaling on domain [0, 255] to increase the contrast of the input image and return the resulting image.

Parameters

input	input bgr or grayscale image.
output	resulting image of autoscaling.

BIMEF() [1/2]

void cv::intensity_transform::BIMEF	(	InputArray	input,
		OutputArray	output,
		float	mu = 0.5f,
		float	a = -0.3293f,
		float	b = 1.1258f
	)

Python:
	output	=	cv.intensity_transform.BIMEF(	input[, output[, mu[, a[, b]]]]	)
	output	=	cv.intensity_transform.BIMEF2(	input, k, mu, a, b[, output]	)

#include <opencv2/intensity_transform.hpp>

Given an input color image, enhance low-light images using the BIMEF method ([276] [277]).

Parameters

input	input color image.
output	resulting image.
mu	enhancement ratio.
a	a-parameter in the Camera Response Function (CRF).
b	b-parameter in the Camera Response Function (CRF).

Warning

This is a C++ implementation of the original MATLAB algorithm. Compared to the original code, this implementation is a little bit slower and does not provide the same results. In particular, quality of the image enhancement is degraded for the bright areas in certain conditions.

BIMEF() [2/2]

void cv::intensity_transform::BIMEF	(	InputArray	input,
		OutputArray	output,
		float	k,
		float	mu,
		float	a,
		float	b
	)

Python:
	output	=	cv.intensity_transform.BIMEF(	input[, output[, mu[, a[, b]]]]	)
	output	=	cv.intensity_transform.BIMEF2(	input, k, mu, a, b[, output]	)

#include <opencv2/intensity_transform.hpp>

Given an input color image, enhance low-light images using the BIMEF method ([276] [277]).

This is an overloaded function with the exposure ratio given as parameter.

Parameters

input	input color image.
output	resulting image.
k	exposure ratio.
mu	enhancement ratio.
a	a-parameter in the Camera Response Function (CRF).
b	b-parameter in the Camera Response Function (CRF).

Warning

This is a C++ implementation of the original MATLAB algorithm. Compared to the original code, this implementation is a little bit slower and does not provide the same results. In particular, quality of the image enhancement is degraded for the bright areas in certain conditions.

contrastStretching()

void cv::intensity_transform::contrastStretching	(	const Mat	input,
		Mat &	output,
		const int	r1,
		const int	s1,
		const int	r2,
		const int	s2
	)

Python:
	None	=	cv.intensity_transform.contrastStretching(	input, output, r1, s1, r2, s2	)

#include <opencv2/intensity_transform.hpp>

Given an input bgr or grayscale image, apply linear contrast stretching on domain [0, 255] and return the resulting image.

Parameters

input	input bgr or grayscale image.
output	resulting image of contrast stretching.
r1	x coordinate of first point (r1, s1) in the transformation function.
s1	y coordinate of first point (r1, s1) in the transformation function.
r2	x coordinate of second point (r2, s2) in the transformation function.
s2	y coordinate of second point (r2, s2) in the transformation function.

gammaCorrection()

void cv::intensity_transform::gammaCorrection	(	const Mat	input,
		Mat &	output,
		const float	gamma
	)

Python:
	None	=	cv.intensity_transform.gammaCorrection(	input, output, gamma	)

#include <opencv2/intensity_transform.hpp>

Given an input bgr or grayscale image and constant gamma, apply power-law transformation, a.k.a. gamma correction to the image on domain [0, 255] and return the resulting image.

Parameters

input	input bgr or grayscale image.
output	resulting image of gamma corrections.
gamma	constant in c*r^gamma where r is pixel value.

logTransform()

void cv::intensity_transform::logTransform	(	const Mat	input,
		Mat &	output
	)

Python:
	None	=	cv.intensity_transform.logTransform(	input, output	)

#include <opencv2/intensity_transform.hpp>

Given an input bgr or grayscale image and constant c, apply log transformation to the image on domain [0, 255] and return the resulting image.

Parameters

input	input bgr or grayscale image.
output	resulting image of log transformations.