Image feature extraction using wavelets

[Ramya D]

hi everyone..
I'm entirely new to matlab and at present I'm working on image processing.
I started with image segmentation and then I wanted to apply a wavelet on my image to extract features.
I want code to apply biorthogonal wavelet on an image to get texture and intensity features from it. Please help.

[Shanmugam Kannappan]

"Ramya D" <ram...@gmail.com> wrote in message <hj150u$slr$1...@fred.mathworks.com>...

Hi,

I am not that femiliar in wavelet but
Try dwt2 & wavmenu in matlab help.

Shan..

[Amir]

Hi,

There exists several method for feature extraction using wavelet transform.
1) First of all you need to apply wavelet transform [cA cH cV cD] =dwt2(I,'haar');

2) Then apply different wavelet features on each sub-band:
a) in a paper entitled "A texture Analysis approach to corrosion image classification", you can extract non-invariant features
b)in another recent paper, "Texture Image Classification Using Visual Perceptual Texture Features and Gabor Wavelet Features" they have introduce 3 new perceptual feature that sounds pretty perfect for segmentation and classification tasks.

[Michael Chan]

Some reference:
http://www.mathworks.com/matlabcentral/fileexchange/29049-wavelet-decomposition-for-images

[Wayne King]

"Ramya D" <ram...@gmail.com> wrote in message <hj150u$slr$1...@fred.mathworks.com>...

Hi Ramya, you can use wavedec2() to decompose your image, enter

waveinfo('bior')

To see the available biorthogonal wavelets supported.

Wayne