How can i differentiate two shape from image?

In my project,I need to identify well shape and deformed shape of fruit..Which technique is helpful?

Answer 1

[Image Analyst]

You could look at the solidity, the major and minor axes lengths, the ratio of the perimeter squared to (4*pi*area). You can get those from regionprops, which I think I showed you how to use in your prior fruit posts. If not, see my image segmentation tutorial http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862

For a more detailed description, see article on "morphometrics" http://en.wikipedia.org/wiki/Morphometrics

[Sabanam]

All this methods are related to Sparse domain.I have all implemented but results are not accurate..What about Discrete wavelet transform.Does it work?

[Image Analyst]

I don't know what the Sparse domain it. And I don't have the wavelet toolbox so I can't answer that question.

[Sabanam]

Ok..

Answer 2

[Karthik Karri]

As the previous answer says, you can use the X and Y axes.Or you can convert the image to gray scale and then do certain comparisons such as the average gray level,histogram,mean, variance etc. Hope my answer helps