remove pixels with certain rgb values

6 Ansichten (letzte 30 Tage)
Tian Tian
Tian Tian am 15 Feb. 2018
Kommentiert: Jona Boender am 28 Apr. 2020
Hi, I want to remove pixels with R value >100 and G value>100, here is my code:
[height width rgb] = size (I); % I is RGB image
for coloum = 1:width
for row =1:height
if (I(row,coloum,1)> 100 && I(row,coloum,2)> 100)
I(row,coloum,1)= 0;
I(row,coloum,2)= 0;
I(row,coloum,3)= 0;
end
end
The error is "Index exceeds matrix dimensions". Does anyone have an idea about this? Thanks in advance.

Akzeptierte Antwort

Image Analyst
Image Analyst am 17 Feb. 2018
Instead of all your code (nested for loops), try this vectorized approach:
mask = (I(:, :, 1) > 100) & (I(:, :, 2) > 100);
% Mask the image using bsxfun() function to multiply the mask by each channel individually.
maskedRgbImage = bsxfun(@times, I, cast(mask, 'like', I));
Also, I very much recommend you don't name your variable I because it looks too much like l (lower case L) and 1 (one).
  5 Kommentare
Image Analyst
Image Analyst am 9 Jul. 2018
In my example, pixels with colors meeting the criteria (red > 100 and green > 200) will be "true" in the mask and therefore will be set to pure black (0,0,0).
A pixel with color of (101, 101, 101) will remain unchanged. Why? It's not in the mask because it's green value was only 101 (not greater than 200). Only the red channel met the criteria but the criteria required that both red and green meet the criteria to be considered as part of the mask.
Jona Boender
Jona Boender am 28 Apr. 2020
This worked great, thank you!

Melden Sie sich an, um zu kommentieren.

Weitere Antworten (1)

Jos (10584)
Jos (10584) am 16 Feb. 2018
Is I really a 3D array or is the third dimension just 1?
Simply take a look at the value of rgb, after the size(I) command :)
  6 Kommentare
Tian Tian
Tian Tian am 17 Feb. 2018
Also, I tried 'colormap', and got the error: 'Colormap must have 3 columns: [R,G,B].'. So it is not an indexed image.
Tian Tian
Tian Tian am 17 Feb. 2018
Thank you I imported a wrong image that's why. Thanks a lot for your help!!

Melden Sie sich an, um zu kommentieren.

Community Treasure Hunt

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

Start Hunting!

Translated by