Clear Filters
Clear Filters

The image is from a leaf, I want to find the discontinuous points ,which is because of the bugbite. How can I find these points as fast as I can?

1 view (last 30 days)
  5 Comments
Mark Sherstan
Mark Sherstan on 13 Dec 2018
Analyze the x and y values you plotted instead of using image processing.If you want to use image processing post the original image.

Sign in to comment.

Answers (2)

Image Analyst
Image Analyst on 13 Dec 2018
You can take the convex hull of the image. Then subtract (use xor function) the initial leaf mask. The difference could be where the bugs have eaten away, assuming the leaves are convex. If you don't know how, then post your original image and the code you used to segment it to a binary image mask.
  3 Comments

Sign in to comment.


Image Analyst
Image Analyst on 13 Dec 2018
Try the attached script. At least it's a start. Adapt as needed.
0000 Screenshot.png
0001 Screenshot.png
  2 Comments
yinghui zhu
yinghui zhu on 13 Dec 2018
Thanks a lot .But as you can see,the simple smooth analysis will make lot of mistakes ,such as the petiole.
Image Analyst
Image Analyst on 14 Dec 2018
I don't know what that is but the algorithm doesn't know if a little incursion/notch in the outline is normal or due to being eaten by an insect. There is no expert knowledge programmed into this simple algorithm.
I suggest you employ the expert knowledge of your user and have them use ginput() to click on regions that you want to remove because they're normal and not an eaten region. Something like
(x, y) = ginput(1); % User clicks on blob to remove.
row = round(y);
column = round(x);
blobLabel = labeledImage(row, column) % Find ID # of blob.
labeledImage(labeledImage == blobLabel) = 0; % Erase that blob.
You can put this into a loop where you ask the user if they need to remove a blob, or accept all existing blobs and continue onwards.

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!