Can anybody help me to explain the logic behind bwdist matlab function. thanks

Question

Muhammad Sohail 2015-11-5

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/253135-can-anybody-help-me-to-explain-the-logic-behind-bwdist-matlab-function-thanks

评论： Guillaume 2015-11-5

Please guide me if someone have the knowledge that how bwdist function work.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Thorsten 2015-11-5

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/253135-can-anybody-help-me-to-explain-the-logic-behind-bwdist-matlab-function-thanks#answer_198561

在 MATLAB Online 中打开

B is a binary image, i.e., an image that contains only 0's and 1's. For every pixel in B, bwdist computes the distance to the nearest 1. So if the pixel is 1, the distance is 0. For a pixel 0 that is e.g., left to a pixel of 1, the distance is 1. And so one. By default, the Euclidean distance is computed, but you can use other distance measures.

 X = zeros(10)
 X(16) = 1
 X(44) = 1
 bwdist(X)

2 个评论
显示无隐藏无

Muhammad Sohail 2015-11-5

Thanks Thorsten for the reply, but actually i am looking for how bwdist measure the distance transform.What are the steps to calculate the distance transform in our own code without using bwdist, for example if use city block distance measure.

Guillaume 2015-11-5

Well, as Thorsten says you go through each pixel (with a for loop for example), find the nearest 1 (you could do that with another for loop, expanding outward from your pixel) and calculate the distance between the two. That would be one (slow!) way of doing it.

Another way of doing it would be to dilate the image by 1 pixel orthogonally each step. At each step, the new pixels reached by the dilation are assigned a distance of 1 + previous step distance. That would only require one loop

请先登录，再进行评论。