Computer Vision analyses videos and images. Images in computer science are visual representations of information in a digital form. To be more specific, an image is a matrix of numbers specifically ordered so that each number represents a part of the information and when put together a visual depiction is created on a screen. Each number in a matrix is represented by a specific pixel on a screen and depending on the value, pixels are light up with various brightness. Since images are matrices their analysis and processing come down to operations on matrices and numbers. For example: mirroring an RGB image is nothing but reversing the rows of the corresponding matrices. The similarity between two images can be recognized if the pixel values are close to each other, or more sophisticated techniques can be used that do not require every pixel to be similar. The same principles apply to videos as well, because they are a sequence of different images called frames that change when a video is played.

More Info