image to gray scale image

How can I convert an image to gray scale image ?

And extract information of that image which is visible in front of light as like government bank note has some hidden feature which is visible in front of light.

This things are required to detect an counterfeit bank note.

please help..............

[Image Analyst]

Duplicate of http://www.mathworks.com/matlabcentral/answers/68779-i-want-to-convert-an-image-to-binary-image-and-then-extract-hidden-information-from-that-image?

[Image Analyst]

Can you post an image and point out what the "hidden" things are, that is, if they're not too hidden and at least somewhat visible if you know where and what to look for? If it's completely 100% hidden then of course it's not there and there's no way to see it.

[Shahed]

yes. you are right.

Now I upload the image please help me to extract those information which specified in picture.

The link of my image is::: http://imageshack.us/photo/my-images/442/10001o.jpg/

Answer 1

[Walter Roberson]

rgb2gray()

for the first part. Feature extraction is going to depend on what you are trying to extract and what the image is like.

[Shahed]

Thank you for your response. Now I upload the image please help me to extract those information which specified in picture.

The link of my image is::: http://imageshack.us/photo/my-images/442/10001o.jpg/

[Image Analyst]

Walter's answer remains the same. That is a complicated image and there are many possible faint patterns to be detected. Each thing you want to detect could be a whole project in itself. This is not some easy thing where we could spend 5 minutes writing code and hand it over to you. You're going to have to do a lot of research into pattern recognition, and standard image processing methods. Start your research here: http://iris.usc.edu/Vision-Notes/bibliography/char1029.html#Money%20and%20Check%20Processing%20--%20Amounts,%20etc.. Lots of papers there on the topic of currency and bank note image processing and authentication.

[Walter Roberson]

You can probably detect the "1000" without a lot of difficulty. The circle around the logo would probably be more difficult. The logo will be tricky to find and verify the shape of, especially if the picture is titled or the illumination is uneven. The flower might be quite difficult to extract and verify: I doubt you would be able to solve that in a reasonable time.