I was working on a fingerprint recognition project. As a part of an orientation field estimation algorithm the input fingerprint image should be divided into blocks and have to compute gradients at each pixel in each block and my question is how this gradients are calculated.
可以将文章内容翻译成中文,广告屏蔽插件可能会导致该功能失效(如失效,请关闭广告屏蔽插件后再试):
问题:
回答1:
You need to be more specific about the type of gradient. Gradient can mean one of:
- Sobel (vertical, horizontal, or combined)
- Laplacian
- Approximations/alternatives to the above (e.g. Scharr)
From looking at your vague question, it sounds like you haven't really tried to understand your problem sufficiently. I suggest that you start by taking a look at the corresponding Wikipedia article, its references and any relevant hits on Google.
Once you've done that, come back and refine your answer -- I'm sure many people will be happy to help you.
回答2:
The answer according to comment info "Compute the gradients Gx and Gy at each pixel in each block" is:
Use the Sobel operators as described on Wikipedia as Gx and Gy made from source A. Then the pixel [x,y] has gradient vector (Gx[x,y], Gy[x,y]).