Mastering Computer Vision with TensorFlow 2.x

Creating an integral image

Integral images are used to calculate rectangular feature pixel values very quickly in one pass. To understand integral images better, let's look at the following breakdown of its calculation:

• The value of the Haar like feature can be calculated as the difference between sum of the pixel values in the white region and the sum of the pixel values in the black region.
• The sum of the pixel I(x,y) can be represented by the value of all of the pixels to the left and above the current pixel location i(x,y), including the current pixel value, and can be represented as follows:

In the following image, I(x,y) is the final ...