How to add box to an image?

58 次查看(过去 30 天)
Hi
I have two images ('yellow.png' and 'bird.png'). See below.
1-4.png bird1_before.jpg
How can I make a box AROUND the image borders with self-defined box colours, say all four lines with grey colour [0.5 0.5 0.5]?
See below.
1-1.png 1-2.png
Please can you help?
  7 个评论
Adam Danz
Adam Danz 2019-8-28
编辑:Adam Danz 2019-8-30
"The images were first generated in Matlab"
Maybe it would be easier if you altered the code that generated the image. Adding the frame is just 1 line of code. But if you have read the images back into matlab, that's going to be more work.
Based on the code and image in your other comment, it appears the image's bottom, left corner is not at (0,0) where your rectangle starts. See answer.
DGM
DGM 2022-7-16
A grayscale border can be added using IPT padarray(). A colored border requires extra steps, though alternative methods exist. See the following reference answer, which shows various possibilities.

请先登录,再进行评论。

采纳的回答

Image Analyst
Image Analyst 2019-8-30
编辑:Image Analyst 2019-8-30
If the bad lines are always the rightmost one and the bottom one, you can simply crop the image to not include those:
rgbImage = rgbImage(1:end-1, 1:end-1, :);
To pad the image with certain color, you can use padarray() on each color channel. There are other ways. Let me know if it's still a problem for you.

更多回答(3 个)

Adam Danz
Adam Danz 2019-8-28
编辑:Adam Danz 2019-9-2
Use the handle output from imshow to get the image's coordinates. The XData and YData properties will provide you with the coordinates of the image. Unless the image coordinates are specified in imshow() inputs, the first pixel is centered at coordinate (1,1) so the rectangle should begin at (0.5, 0.5). To expand the rectangle to the outer edges, you must add 1 to the width and height indicated by ranges of XData and YData.
% When the image is in pixel units
corn_gray = imread('corn.tif',3);
h = imshow(corn_gray);
rh = rectangle('position',[ ...
h.XData(1)-.5,...
h.YData(1)-.5,...
range(h.XData)+1,...
range(h.YData)+1],...
'EdgeColor', 'r','linewidth',2);
When the image is not in pixel units, XData and YData properties of the image object may no longer indicate the pixel centers so you must compute the location of the pixel edges.
% When the image is not in pixel units
corn_gray = imread('corn.tif',3);
h = imshow(corn_gray,'XData',200:400,'YData',110:4000); %(arbitrary xdata,ydata)
axis square
axis on % optional, to see axis scales
pixelRatio = [range(h.XData)/size(corn_gray,2), range(h.YData)/size(corn_gray,1)];
rh = rectangle('position',[ ...
h.XData(1)-pixelRatio(1)/2,...
h.YData(1)-pixelRatio(2)/2,...
range(h.XData)+pixelRatio(1),...
range(h.YData)+pixelRatio(2)],...
'EdgeColor', 'r','linewidth',2);
% Expand the axis limits to show full frame
xlim(xlim()+[-.01,.01]*range(xlim())) % add 1% to the x axis limits
ylim(ylim()+[-.01,.01]*range(ylim())) % add 1% to the y axis limits

Image Analyst
Image Analyst 2019-8-31
Images rows and columns start with 1, not zero. so do this:
rectangle('position', [1, 1, s(2), s(1)], 'edgecolor', [1 0 0])
If some part of the rectangle is still missing (outside the image), try adding or subtracting 0.5 or 1 to the position values until you can see it. Like perhaps
rectangle('position', [1, 1, s(2) - 1, s(1) - 1], 'edgecolor', [1 0 0])
  5 个评论
Adam Danz
Adam Danz 2019-9-3
编辑:Adam Danz 2019-9-3
Then you should have accepted Image Analyst's answer rather than your modification of another person's work.
Salad Box
Salad Box 2020-1-30
I have accepted Image Analyst's anwser.

请先登录,再进行评论。


Salad Box
Salad Box 2019-9-3
clear
close all
clc
RGB = imread('bird1.jpg');
s = size(RGB);
figure
imshow(RGB);
rectangle('position',[1 1 s(2) s(1)], 'edgecolor', [1 0 0])
rectangle('position',[0 0 s(2) s(1)], 'edgecolor', [1 0 0])
3.png
  1 个评论
Adam Danz
Adam Danz 2019-9-3
编辑:Adam Danz 2019-9-6
If you zoom into the image at the rectangle corners you'll see that it doesn't frame the pixel edges because you're not offsetting it by the pixel width & height.

请先登录,再进行评论。

Community Treasure Hunt

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

Start Hunting!

Translated by