sameerr11
Junior Member level 1
"An efficient watershed algorithm based on connected components" by A. Bieniek . I have implemented this algorithm in matlab. I am new to image processing field.
One image example is given in paper .It is mentioned that taking a gradient image thresholded with an arbitrary value then watershed segmentation algorithm applied. I used sobel filter to find gradient of each pixel, which give dx and dy component for each pixel. Then i take magnitude of it by sqrt(dx^2 + dy^2).
After that i use arbitary value to threshold gradient magnitude , i make gradient mag to zero for all (gradmag<threshold) .Is it correct way ?
I have final matrix output with correct labeling. But i dont get one thing, how do theyget binary segmented image(as you showed in paper in Fig3). I have labeled output image which has intensity value of label for all pixels. Is there any final stage which i need to perform on image to get binary value.
One image example is given in paper .It is mentioned that taking a gradient image thresholded with an arbitrary value then watershed segmentation algorithm applied. I used sobel filter to find gradient of each pixel, which give dx and dy component for each pixel. Then i take magnitude of it by sqrt(dx^2 + dy^2).
After that i use arbitary value to threshold gradient magnitude , i make gradient mag to zero for all (gradmag<threshold) .Is it correct way ?
I have final matrix output with correct labeling. But i dont get one thing, how do theyget binary segmented image(as you showed in paper in Fig3). I have labeled output image which has intensity value of label for all pixels. Is there any final stage which i need to perform on image to get binary value.
