How To Recognize Different Object Types

I am trying to recognize different cell types which are oxalate cyristals, red blood cells, epithelial in urine sediment images. For now I used hough transform for red blood cell, morphological filters for oxalate cyristals and epithelial cells. But my problem is; if there is more than one cell type in image, my code does not work. How can I solve this problem? Thanks for your helps.

[MFK]

<<

>>

[MFK]

[MFK]

Both images are same. But as you can see, in first image I used only detection epithelial code, in second image I used detection red blood cell, epithelial cells and oxalate cyristals codes. How can i solve this problem?

Answer 1

[Image Analyst]

This is more of an algorithm development question, rather than a MATLAB programming question, even though you're going to use the MATLAB language to program it up. There is no built in code to detect "red blood cell, epithelial cells and oxalate cyristals codes". So first you'll have to develop an algorithm to detect each of those. Once you have that, you can OR them all together to get a binary image with all of them in there. Maybe try the newsgroup sci.image.processing if you want algorithm advice, or check out Vision Bibliography.