What is Convolution in Computer Vision?

Question

Could someone tell me what is a Convolution in Computer Vision?

Answer 1

Convolution is basically a filter that could be applied to images from the Computer Vision’s perspective. Convolution in mathematical terms, is defined as an operator, which takes in two functions, and then produces a third function which best represents the amount of overlap between those two functions, but in a very different manner. Now you may be wondering how it works, and how it could apply a filter to an image. Convolution works by determining the value of the central pixel, by summing up the weighted values of all of the neighbor points together, and then creates a modified filtered image.

If you wish to get certified, in Computer Vision, then have a look at Intellipaat’s Artificial Intelligence course, which will offer you industry-grade course with guided projects and exercises to help you enhance and amp up your practical skills and experience in the domain of Artificial Intelligence.

Also, if you are a novice in the field of Computer Vision, have a look at the following video on Introduction to Computer Vision and get a better understanding.