Computer Vision is perhaps the most intriguing and fascinating concept in artificial intelligence. Computer Vision is an interdisciplinary field that deals with how computers or any software can learn a high-level understanding of the visualizations in the surroundings. After obtaining this conceptual perspective, it can be useful to automate tasks or perform the desired action.

The tasks that are obvious to the human brain are not so intuitive to the computers as they need to be trained specifically on these jobs to produce effective results. This process involves complicated steps like acquiring the data from the real world, processing the acquired data in a suitable format, analyzing the processed images, and finally teaching and training the model to perform the complex task with very high accuracy.

To understand computer vision more intuitively, let us consider an example. Assume you have to teach a computer to differentiate between the various colors. Consider, you have three objects with the following colors, namely red, blue, and green, and you want to differentiate these colors accordingly. This job is an extremely simple task for the human brain to perform, but it is quite a complicated task for the computer to perform.

The task mentioned above is one of the most basic actions that can be performed by using computer vision. We will learn about how images in the digital world work and also try to understand the image and how these stacked layers exactly work. . We will also learn in-depth about the basics of the open-cv module. Finally, we will also be implementing some hands on basic level projects with this library. So, without further ado let us dive into understanding all the aspects required for mastering the basic computer vision skills.

Dealing With Images:

Screenshot By Author

The composition of these three colors, namely red, green, and blue can be used to compose almost any other color. Mixing them in the right proportion allows us to frame any other desired color. This concept has existed since the cathode ray televisions a few decades ago. So how does this exactly work?

Each of these colors has an 8 bit integer value. This means a matrix of these could range from 0 to 255. The reasoning for this is because 2⁸ is 256 and 0–255 consist of 256 values. Each of these colors will have a value of this range and since we have a 3-Dimensional image, we can stack each of these upon each other. This might be a slightly more complex example, so let us switch over to the grayscale images which only consists of black and white and that would be easier to understand. Below is the grayscale representation.

Screenshot By Author

This grayscale representation shown above should be a good starting point to understand the concept of how images work in computer vision better. The below figure shows how the level change starts to happen as we move from the 0th mark to the 255th mark. After 256 levels of changing we go from a completely black shade to a fully white shade.

#programming #machine-learning #computer-vision #data-science #opencv