How to detect inclined lines in an image?

i have a question regarding line detection. I have an image (in fact edge image/ edge map) containing some horizontal, some vertical and some inclined lines. I want to find only inclined lines from that image. How can i do that by searching the entire edge map, as i don't have any information about these lines? I have to search the entire map to search the coordinates (or an开发者_如何学Goy other related information). Now can any one help me out how to do that?

Thanks alot

---

the first thought i have about this problem is only a varient the way to use the hough transform because you can analize the acumulator of hough transform, in this acumulator you have the angle and rho of every line detected you can define that only you care the line that have a angle diferent to 0 and 90

http://homepages.inf.ed.ac.uk/amos/hough.html

http://www.cs.sfu.ca/~hamarneh/ecopy/compvis1999_hough.pdf

Is the best way i can imagine because you have the angle of each line

---

This is a maths question really. What you want is a Radon transform: http://en.wikipedia.org/wiki/Radon_transform

---

I recently had to write a sphere detection algorithm, which was also based on edge detection. You state that you know nothing about the inclined lines. This is problematic. You could for example filter out whatever edge points belong to staight lines and then use an algorithm to group the points which belong to inclined lines. You could then use linear equations to find the characteristics of the inclined lines.

Hope it helped.