In the previous blog, we discussed the Bayer filter and how we can form a color image from a Bayer image. But we didn’t discuss much about interpolation or demosaicing algorithms so in this blog let’s discuss these algorithms in detail.
According to Wikipedia, Interpolation is a method of constructing new data points within the range of a discrete set of known data points. Image interpolation refers to the “guess” of intensity values at missing locations.
The big question is why we need interpolation if we are able to capture intensity values at all the pixels using Image sensor?
- Bayer filter, where we need to find missing color information at each pixel.
- Projecting low-resolution image to a high-resolution screen or vice versa. For example, we prefer watching videos in the full-screen mode.
- Image Inpainting, Image Warping etc.
- Geometric Transformations.
There are plenty of Interpolation methods available but we will discuss only the frequently used. Interpolation algorithms can be classified as
Non-adaptive perform interpolation in a fixed pattern for every pixel, while adaptive algorithms detect local spatial features, like edges, of the pixel neighborhood and make effective choices depending on the algorithm.
Let’s discuss the maths behind each interpolation method in the subsequent blogs.
In the next blog, we will see how the nearest neighbor method works. Hope you enjoy reading.
If you have any doubt/suggestion please feel free to ask and I will do my best to help or improve myself. Good-bye until next time.