Image restoration : Noise Removal Techniques in Image Preprocessing

Noises introduces unwanted effects over image, due to various reasons and they can be approximate by using probability density function, for more detail on noises in image checkout previous post on image restoration ( https://sumitkrsharma-ai.medium.com image-restoration-types-of-noise-in-image-preprocessing-8434009278b9 ) which includes types of noise in image pre-processing briefly .

This post is about exploring ways to handle various different types of noises in image preprocessing. Let’s start with boarder classification of filters that are capable to handle noises.

Types of filters

• Spatial Filters

As name suggests these filters relies over the spatial information it means that they handle the noise by manipulating pixel values based on neighboring pixels. The the kernels get the information that how much the output is influenced with the neighborhood pixel values.

Examples of noises handled by these filters — Gaussian noise, salt and pepper, and speckle noise.

• Frequency Filters

They works upon the spatial frequrency component of the image and improves the image by manupulating them based on the type of noises that impacted the image, where spatial frequency…