Image smoothing (smooth image)

May 27 • Notes • 1184 Views • 1 Comment on Image smoothing (smooth image)

  image smoothing (Smooth image ):

Photos comprise rich and well-structured visual information. In human visual perception, edges are effective and expressive stimulation, vital for neural inter

presentation to make the best sense of the scene. In manipulating and understanding pictures, high-level inference with regard to salient structures was intensively attended to. Research following this line embodies generality and usefulness in a wide range of applications, including image recognition, object classification,  segmentation, and many other smooth image and non-photo realistic rendering tasks.

 

image smoothing Smooth image

image smoothing Smooth image

Notation :

Image = 2D array of pixels

Pixel = intensity (scalar) or color (3D vector)

Ip = value of image I at position: p = ( px , py )

F [ I ] = output of filter F applied to image I

 

 

in this paper present a new smooth image tool, helpful

for  enhancing and characterizing  fundamental image constituents,

i.e., salient edges, and in the meantime for diminishing insignificant

details.  Method relates in spirit to edge-preserving

smoothing  [ Tomasi and Manduchi 1998; Durand and Dorsey 2002;

Paris and Durand 2006;  Farbman et al. 2008; Sub retal. 2009;

Kass and Solomon 2010] that aims to retain primary color change,

and yet it  differs from them  in focus ,in essence and in mechanism.

Our objective is to globally maintain and possibly enhance the most

prominent set of edges by increasing steepness of transition while

not affecting the overall acutance. It enables faithful principal structure

representation.

 

Strategy for Smoothing Images

 

Images are not smooth because adjacent pixels are different.

Smoothing = making adjacent pixels look more similar.

Smoothing strategy pixel -> average of its neighbors.

Algorithmic-ally, we propose a sparse gradient counting scheme in

an optimization framework. The main contribution is a new strategy

to confine the discrete number of intensity changes among neighboring

pixels, which links mathematically to the L0norm for information

sparsity pursuit. This idea also leads to an unconventional

global optimization procedure involving a discrete metric,

which solution enables diversified of the edge manipulation according to

saliency. The qualitative effect of our method is to thin salient

edges, which makes them easier to be detected and more visually

distinct. Different from color quantization and segmentation, our

enhanced edges are generally in line with that original . Even

small objects resolution and thin edges may be faithfully maintained

if they are conspicuous in structurally . The framework is general and finds so many applications. We can  apply

it to compression artefact degraded  recovery of clip-art. High quality results can be obtained in the experiments which are extensive . Our method can also profit edge extraction, a fundamentally important operator,

by effectively removing noise part , even of slight blurriness,  and unimportant details , making the results immediately usable in image abstraction and pencil sketch production. In traditional decomposition of layer , with an additional step to

avoid structure over-enhancement, our method is applicable to detail enhancement based on possibly to HDR tone mapping , and separating layers  after parameter tuning. We show several examples along with discussion of limitations that our method might cause over-sharpening when strong smoothing is applied  for large illumination variation spanning dozens of pixels .

 

1D Smoothing:

We  enhance edges of highest-contrast  by confining the number of nonzero

gradients, if smooth image is achieved in a global manner.

To begin with, we denote the input discrete signal by g and its

smoothed result by f . Our method counts amplitude changes discretely

2D Formulation:

In 2D image representation, we denote by I the input image and by

S the computed result. The gradient ∇Sp = (∂xSp, ∂ySp)T for each

pixel p is calculated as colour difference between neighbouring pixels

along the x and y directions

Q1: What is 1D image formulation?

Ans: We  enhance edges of highest-contrast  by confining the number of nonzero

gradients, if smooth image is achieved in a global manner.

To begin with, we denote the input discrete signal by g and its

smoothed result by f . Our method counts amplitude changes discretely called 1D smoothing.

Q2: What is 2D image formulation ?

Ans: In 2D image representation, we denote by I the input image and by

S the computed result. The gradient ∇Sp = (∂xSp, ∂ySp)T for each

pixel p is calculated as colour difference between neighbouring pixels

along the x and y directions called 2d formulation.

Q3: how many types of imageformulation are there?

Ans: 2 types,1D and 2D formulation.

Q4: Image is equal to how much pixels?

Ans: Image = 2D array of pixels

Q5: Pixel is equal to what parameter?

Ans: Pixel = intensity (scalar) or color (3D vector).

Tell us Your Queries, Suggestions and Feedback

Your email address will not be published.

One Response to Image smoothing (smooth image)

  1. Prabhat Saxena says:

    image smoothing is a technique of image processing .this article is about image processing that is to smooth a data set is to create an approximating function that attempts to capture important patterns in the data, while leaving out noise or other fine-scale structures/rapid phenomena.

« »