Automating cropping of an image.
in [Matlab]
|
From: Alex Jin on 20 Apr 2010 11:55 ImageAnalyst <imageanalyst(a)mailinator.com> wrote in message <f81083ce-09e9-4bb9-9d8b-03df61595641(a)30g2000yqi.googlegroups.com>... > Well if you want to find the first and last rows (or columns) that > have a 1 in them, don't you need to check which rows have a 1 in > them? Think about it. How would you do it? If you can't figure it > out then study the demo below: > > clc; > % Read in standard MATLAB grayscale demo image. > grayImage = double(imread('moon.tif')); > subplot(2, 2, 1); > imshow(grayImage, []); > title('Original Grayscale Image'); > set(gcf, 'Position', get(0,'Screensize')); % Maximize figure. > > binaryImage = grayImage > 100; > subplot(2, 2, 2); > imshow(binaryImage, []); > title('Binary Image'); > col1 = find(any(binaryImage, 1), 1, 'first') > col2 = find(any(binaryImage, 1), 1, 'last') > row1 = find(any(binaryImage, 2), 1, 'first') > row2 = find(any(binaryImage, 2), 1, 'last') > boxX = [col1 col2 col2 col1 col1]; > boxY = [row1 row1 row2 row2 row1]; > hold on > plot(boxX, boxY, 'r-', 'LineWidth', 2); > > croppedImage1 = imcrop(grayImage, [col1, row1, (col2-col1+1) (row2- > row1+1)]); > subplot(2, 2, 3); > imshow(croppedImage1, []); > title('Cropped Gray Image'); > > croppedImage2 = imcrop(binaryImage, [col1, row1, (col2-col1+1) (row2- > row1+1)]); > subplot(2, 2, 4); > imshow(croppedImage2, []); > title('Cropped Binary Image'); Thank you so much.. i understand more.. but by using this way i cant really crop out the wanted area of my image.. This link shows an example image i would like crop. the area i want is in red: http://www.postimage.org/image.php?v=TskFOOJ As i can see, the area i wanted to crop has a rougher surface, anyway i can use this to crop out the image? thanks
From: Alex Jin on 20 Apr 2010 18:42 ImageAnalyst <imageanalyst(a)mailinator.com> wrote in message <d4a50311-409e-4328-a719-6eae29475f2b(a)y3g2000vbg.googlegroups.com>... > Cropping is not your problem. You aren't even close to segmenting > your region of interest. That is too complicated an image to do > simple thresholding on. You need to figure out how to get that first, > before you start thinking about cropping. Get the values across the middle of the image and high pass or band pass it..? Do u have any example code for band pass and thresholding operation? thanks
From: Alex Jin on 21 Apr 2010 09:42
ImageAnalyst <imageanalyst(a)mailinator.com> wrote in message <d4a50311-409e-4328-a719-6eae29475f2b(a)y3g2000vbg.googlegroups.com>... > Cropping is not your problem. You aren't even close to segmenting > your region of interest. That is too complicated an image to do > simple thresholding on. You need to figure out how to get that first, > before you start thinking about cropping. i had covert the image to greyscale and then plot(Image(200,:)) gives a graph.. i would like to use a high pass filter.. any example codes for that? thanks |