Autobrightness

This post is about an application that I tried to make using the OpenCV Image Processing Library. This application can detect the changes in brightness of the environment and adjust your laptop’s screen brightness accordingly.

This post was published in 2014 and as such the program might not be up to date and as a result may not work properly. However it was a cool project at the time!!

Pixel is the smallest element in an image with definite properties like color, position, intensity etc. A typical colored pixel has three channels RED, GREEN and BLUE. Let me denote this as a 3-tuple \((R,G,B)\). Colors like magenta, yellow etc. are derived from these 3 primary colors and hence they are called derived colors. In image processing the derived colors are the various combinations of \((R,G,B)\). For example If I consider the RED color as \((255,0,0)\), BLUE as $(0,0,255)$ and GREEN as \((0,255,0)\). Then YELLOW can be represented as \((255,255,0)\).

Gray-scale images are those images which has only one channel and let us denote this by \(I\). Grayscale images are useful when it comes to processing images, as it reduces three channels to one and thus eliminates redundancy. This is an example of a typical Grayscale image.

Now after understanding what colored images and grayscale images are; let us move towards the basic algorithm of the application.

First, the application captures images from the laptop’s webcam. Second, the application computes the average intensity (I) of the entire camera input. Changes brightness according to the intensity values. As an extra feature the application also supports Gesture Brightness Control. Now, the question is how to get the average intensity?

Mat img;
int s = (int)img.at<uchar>(j,i);

This is an overloaded C++ template function which returns the pixel value at the position (j,i). This function is only used for a grayscale image i.e. ‘img’ should be grayscale image.

To get the average intensity of all pixels in a grayscale image, we can use a ‘for’ loop.

unsigned long int sum;
for(int i=0; i<img.rows; i++)
{
    for(int j=0 j<img.cols; j++)
    {
        sum += (int)img.at<uchar>(i,j);
    }
}
int intensity = sum/(img.rows*img.cols);

Now we have computed the average intensity of the frame. Now our next step is to change the brightness depending on the value of the intensity. In linux, the brightness can be changed by editing.

/(file system) -> sys -> class -> backlight -> acpi_video0 -> brightness

To change the brightness, open the terminal and run this command. (As root)

echo 10 > /sys/class/backlight/acpi_video0/brightness

If you want to try out the application, download the source tar-ball from here.

Then extract the archive, run the install.sh script.

tar -xvf Autobrightness.tar.gz
cd Autobrightness
chmod +x install.sh
./install.sh

After installing, close and reopen the terminal and type

autobrightness

to start the application.

This installation script is only for linux (Ubuntu/mint), but the source code can be compiled in other Operating Systems like Windows etc.