Introduction to Image Smoothing:

So in today’s blog, I will explain what is Image smoothing, why we need it and how it works. I am going to use the method published by Springerlink.com Authors are : Quian Liu, Caiming Zhang, Qiang Guo and Yuanfeng Zhou. According to the article: The main aim of image in digital image processing is to smooth image. The smoothing process has been widely used in many fields, such as image display, image transmission and image analysis, etc. Image smoothing has been a basic module in almost all the image processing systems. Image smoothing is a method of improving the quality of images. The objects processed are images that are corrupted by different factors during the course of generating, transmitting, processing and displaying loaded image. The whole image smoothing process is like a special part of technology that does the process of image recovery.

  The first smoothing procedure is called the structural smoothing. Natural images contain both the clear structural edges of objects and abundant details caused by lightness, textures, and so on. Psychological studies show that 'human beings tend to pay more attention to the outlines of objects and abundant details caused by lightness, textures and so on' According to this study, Human beings tend to pay more attention to the outlines of objects than trivial details. Indeed, images containing main structures but without details can be of use in many applications such as edge extraction, image abstraction, and tone mapping. Smoothing image aims to produce images which discard insignificant details while preserving the main structural edges. Because of the complexity of natural images, there is no perfect explanation to the computer to differentiate between edges and trivial details: human eyes can make flexible decisions quickly and easily.

The next process is Noise reduction. According to the article by SciVerse science direct ‘Noise may be arisen in the capturing and transmission process of the image.’ The Noise are divided into categories of Gaussian noise, Balanced noise and impulse noise. The Impulse noises displays light and dark noise pixels under random distribution on the image. Thus, impulse noises has important significant to image processing. The following images shows Noises in image:

                                            Noise in gray scale

Noise in color

The third smoothing we are implementing in the project is Gaussian Smoothing, The Gaussian Smoothing operator is a 2-D Convolution Operator, which means that it provides, a way of ‘multiplication together’ two arrays of numbers that is used to ‘blur’ images and remove detail and noise. Gaussian Smoothing is simpler in mean filter, which is done by reducing the amount of intensity variation between one pixel and the next. It is often used to reduce noise in images and get closer to the objects in the image. The problem we might face using Gaussian blur is that it uses different kernel that represent the shape of the a Gaussian hump. Kernel is a smallish matrix that is used in image convolution, which is also referred as ‘Structuring Element’. Here is the formula for Gaussian Distribution: