3

I am using the vl_phow function (http://www.vlfeat.org/matlab/vl_phow.html) and wondering why the smoothing is applied before the feature extraction.

To be more specific in the documentation of vl_phow it is mentioned that:

The image is smoothed by a Gaussian kernel of standard deviation SIZE / MAGNIF. Note that, in the standard SIFT descriptor, the magnification value is 3; here the default one is 6 as it seems to perform better in applications.

So why is this smoothing operation is done?

Also in the same documentation there is WindowSize option which is explained as follows: size of the Gaussian window in units of spatial bins. Is this Window used to smooth the image or for something else?

Can you please tell me why it is done and the advantages of smoothing the image before the feature extraction.

Ander Biguri
  • 35,140
  • 11
  • 74
  • 120
burak akdemir
  • 153
  • 10

3 Answers3

4

While this is more a math question than a MATLAB question , I'll go for it anyway.

SIFT features are supposed to be points that "stand out" in the image. They are features that have high information content, and that will be "invariant" in different images.

However, a noisy image may have "noise" that looks like something important. Basic example:

[0 0 0 0 0
 0 0 0 2 2
 0 1 0 2 2
 0 0 0 0 0]

Without smoothing, one may think that there are 2 areas with loads of information there, the area with 2s and the area with a single 1. However, while the area with 2s does seem like part of the information, as there are a lot of them together, the part with a single 1 may be just noise, a small random value added there due to the noise in the imaging technology.

If you smooth the image with a filter you'dd get something like (made up example)

[0 0    0 0   0
 0 0    0 1.9 2
 0 0.01 0 1.9 2
 0 0    0  0  0]

Where it is way more obvious that the 1 is just noise, but the 2s stay.

That;s why feature extraction algorithms, such as SIFT, do generally smooth the image before getting keypoints.

The bigger the smoothing window, the more robust keypoints you'll find, as smaller things will be removed. However you will also find less keypoints. Also making it too big you risk deleting real information (the 2s int he example).

Ander Biguri
  • 35,140
  • 11
  • 74
  • 120
  • I see your point but smoothing is done in the vl_phow function which extracts the features like dense sift. In other words, it does NOT look for the key points. It extracts the features from regularly distributed points in the image. So there must be another purpose rather than eliminating the noise. – burak akdemir May 11 '16 at 13:20
  • @burakakdemir consider accepting the answer if it helped you – Ander Biguri Jul 25 '16 at 11:51
0

To remove noise in the image. Features (traditionally corner points in intensity) can be greatly affected by noise.

John Wiggles
  • 36
  • 2
0

No camera is free of noise. Unless you have an artificial image you will always have noise in your image.

This noise has many sources, changes over time, temperature and even location.

You will find that applying a lowpass filter like Gaussian blur to the input image or averaging over multiple shots is one of the first steps in pretty much any image processing operation.

Without that it would be nearly impossible to get the same results for two images from the same unchanged scene.

For a quick overview of noise sources and denoising methods read: http://www.ijaiem.org/volume2issue11/IJAIEM-2013-11-24-070.pdf

Piglet
  • 27,501
  • 3
  • 20
  • 43