1 Changing the contrast and brightness of an image! {#tutorial_basic_linear_transform}
2 =================================================
7 In this tutorial you will learn how to:
10 - Initialize a matrix with zeros
11 - Learn what @ref cv::saturate_cast does and why it is useful
12 - Get some cool info about pixel transformations
13 - Improve the brightness of an image on a practical example
19 The explanation below belongs to the book [Computer Vision: Algorithms and
20 Applications](http://szeliski.org/Book/) by Richard Szeliski
24 - A general image processing operator is a function that takes one or more input images and
25 produces an output image.
26 - Image transforms can be seen as:
27 - Point operators (pixel transforms)
28 - Neighborhood (area-based) operators
32 - In this kind of image processing transform, each output pixel's value depends on only the
33 corresponding input pixel value (plus, potentially, some globally collected information or
35 - Examples of such operators include *brightness and contrast adjustments* as well as color
36 correction and transformations.
38 ### Brightness and contrast adjustments
40 - Two commonly used point processes are *multiplication* and *addition* with a constant:
42 \f[g(x) = \alpha f(x) + \beta\f]
44 - The parameters \f$\alpha > 0\f$ and \f$\beta\f$ are often called the *gain* and *bias* parameters;
45 sometimes these parameters are said to control *contrast* and *brightness* respectively.
46 - You can think of \f$f(x)\f$ as the source image pixels and \f$g(x)\f$ as the output image pixels. Then,
47 more conveniently we can write the expression as:
49 \f[g(i,j) = \alpha \cdot f(i,j) + \beta\f]
51 where \f$i\f$ and \f$j\f$ indicates that the pixel is located in the *i-th* row and *j-th* column.
56 - The following code performs the operation \f$g(i,j) = \alpha \cdot f(i,j) + \beta\f$ :
57 @include BasicLinearTransforms.cpp
62 -# We begin by creating parameters to save \f$\alpha\f$ and \f$\beta\f$ to be entered by the user:
63 @snippet BasicLinearTransforms.cpp basic-linear-transform-parameters
65 -# We load an image using @ref cv::imread and save it in a Mat object:
66 @snippet BasicLinearTransforms.cpp basic-linear-transform-load
67 -# Now, since we will make some transformations to this image, we need a new Mat object to store
68 it. Also, we want this to have the following features:
70 - Initial pixel values equal to zero
71 - Same size and type as the original image
72 @snippet BasicLinearTransforms.cpp basic-linear-transform-output
73 We observe that @ref cv::Mat::zeros returns a Matlab-style zero initializer based on
74 *image.size()* and *image.type()*
76 -# Now, to perform the operation \f$g(i,j) = \alpha \cdot f(i,j) + \beta\f$ we will access to each
77 pixel in image. Since we are operating with BGR images, we will have three values per pixel (B,
78 G and R), so we will also access them separately. Here is the piece of code:
79 @snippet BasicLinearTransforms.cpp basic-linear-transform-operation
81 - To access each pixel in the images we are using this syntax: *image.at\<Vec3b\>(y,x)[c]*
82 where *y* is the row, *x* is the column and *c* is R, G or B (0, 1 or 2).
83 - Since the operation \f$\alpha \cdot p(i,j) + \beta\f$ can give values out of range or not
84 integers (if \f$\alpha\f$ is float), we use cv::saturate_cast to make sure the
87 -# Finally, we create windows and show the images, the usual way.
88 @snippet BasicLinearTransforms.cpp basic-linear-transform-display
91 Instead of using the **for** loops to access each pixel, we could have simply used this command:
93 image.convertTo(new_image, -1, alpha, beta);
95 where @ref cv::Mat::convertTo would effectively perform *new_image = a*image + beta\*. However, we
96 wanted to show you how to access each pixel. In any case, both methods give the same result but
97 convertTo is more optimized and works a lot faster.
102 - Running our code and using \f$\alpha = 2.2\f$ and \f$\beta = 50\f$
104 $ ./BasicLinearTransforms lena.jpg
105 Basic Linear Transforms
106 -------------------------
107 * Enter the alpha value [1.0-3.0]: 2.2
108 * Enter the beta value [0-100]: 50
113 ![](images/Basic_Linear_Transform_Tutorial_Result_big.jpg)
118 In this paragraph, we will put into practice what we have learned to correct an underexposed image by adjusting the brightness
119 and the contrast of the image. We will also see another technique to correct the brightness of an image called
122 ### Brightness and contrast adjustments
124 Increasing (/ decreasing) the \f$\beta\f$ value will add (/ subtract) a constant value to every pixel. Pixel values outside of the [0 ; 255]
125 range will be saturated (i.e. a pixel value higher (/ lesser) than 255 (/ 0) will be clamp to 255 (/ 0)).
127 ![In light gray, histogram of the original image, in dark gray when brightness = 80 in Gimp](images/Basic_Linear_Transform_Tutorial_hist_beta.png)
129 The histogram represents for each color level the number of pixels with that color level. A dark image will have many pixels with
130 low color value and thus the histogram will present a peak in his left part. When adding a constant bias, the histogram is shifted to the
131 right as we have added a constant bias to all the pixels.
133 The \f$\alpha\f$ parameter will modify how the levels spread. If \f$ \alpha < 1 \f$, the color levels will be compressed and the result
134 will be an image with less contrast.
136 ![In light gray, histogram of the original image, in dark gray when contrast < 0 in Gimp](images/Basic_Linear_Transform_Tutorial_hist_alpha.png)
138 Note that these histograms have been obtained using the Brightness-Contrast tool in the Gimp software. The brightness tool should be
139 identical to the \f$\beta\f$ bias parameters but the contrast tool seems to differ to the \f$\alpha\f$ gain where the output range
140 seems to be centered with Gimp (as you can notice in the previous histogram).
142 It can occur that playing with the \f$\beta\f$ bias will improve the brightness but in the same time the image will appear with a
143 slight veil as the contrast will be reduced. The \f$\alpha\f$ gain can be used to diminue this effect but due to the saturation,
144 we will lose some details in the original bright regions.
148 [Gamma correction](https://en.wikipedia.org/wiki/Gamma_correction) can be used to correct the brightness of an image by using a non
149 linear transformation between the input values and the mapped output values:
151 \f[O = \left( \frac{I}{255} \right)^{\gamma} \times 255\f]
153 As this relation is non linear, the effect will not be the same for all the pixels and will depend to their original value.
155 ![Plot for different values of gamma](images/Basic_Linear_Transform_Tutorial_gamma.png)
157 When \f$ \gamma < 1 \f$, the original dark regions will be brighter and the histogram will be shifted to the right whereas it will
158 be the opposite with \f$ \gamma > 1 \f$.
160 ### Correct an underexposed image
162 The following image has been corrected with: \f$ \alpha = 1.3 \f$ and \f$ \beta = 40 \f$.
164 ![By Visem (Own work) [CC BY-SA 3.0], via Wikimedia Commons](images/Basic_Linear_Transform_Tutorial_linear_transform_correction.jpg)
166 The overall brightness has been improved but you can notice that the clouds are now greatly saturated due to the numerical saturation
167 of the implementation used ([highlight clipping](https://en.wikipedia.org/wiki/Clipping_(photography)) in photography).
169 The following image has been corrected with: \f$ \gamma = 0.4 \f$.
171 ![By Visem (Own work) [CC BY-SA 3.0], via Wikimedia Commons](images/Basic_Linear_Transform_Tutorial_gamma_correction.jpg)
173 The gamma correction should tend to add less saturation effect as the mapping is non linear and there is no numerical saturation possible as in the previous method.
175 ![Left: histogram after alpha, beta correction ; Center: histogram of the original image ; Right: histogram after the gamma correction](images/Basic_Linear_Transform_Tutorial_histogram_compare.png)
177 The previous figure compares the histograms for the three images (the y-ranges are not the same between the three histograms).
178 You can notice that most of the pixel values are in the lower part of the histogram for the original image. After \f$ \alpha \f$,
179 \f$ \beta \f$ correction, we can observe a big peak at 255 due to the saturation as well as a shift in the right.
180 After gamma correction, the histogram is shifted to the right but the pixels in the dark regions are more shifted
181 (see the gamma curves [figure](Basic_Linear_Transform_Tutorial_gamma.png)) than those in the bright regions.
183 In this tutorial, you have seen two simple methods to adjust the contrast and the brightness of an image. **They are basic techniques
184 and are not intended to be used as a replacement of a raster graphics editor!**
188 Code for the tutorial is [here](https://github.com/opencv/opencv/blob/3.4/samples/cpp/tutorial_code/ImgProc/changing_contrast_brightness_image/changing_contrast_brightness_image.cpp).
189 Code for the gamma correction:
191 @snippet changing_contrast_brightness_image.cpp changing-contrast-brightness-gamma-correction
193 A look-up table is used to improve the performance of the computation as only 256 values needs to be calculated once.
195 ### Additional resources
197 - [Gamma correction in graphics rendering](https://learnopengl.com/#!Advanced-Lighting/Gamma-Correction)
198 - [Gamma correction and images displayed on CRT monitors](http://www.graphics.cornell.edu/~westin/gamma/gamma.html)
199 - [Digital exposure techniques](http://www.cambridgeincolour.com/tutorials/digital-exposure-techniques.htm)