Comments on: Gray scale pixel values in labeled regions

By: Steve

Steve — Tue, 06 Oct 2009 11:16:04 +0000

Opcs—Here's basic assignment:

% Replace element in row 5, col 6 with value 10
A(5, 6) = 10;

You might also want to look at my posts on logical and linear indexing.

By: opcs

opcs — Mon, 05 Oct 2009 21:11:25 +0000

sir,i have an image and want to perform some operations on selective pixels of the image, means to transform an image into different image.

how to apply that function on that pixels, leaving remaining pixels unaffected in the resulted image ?

I’m in new in matlab and have no more time to complete the assignment so please help me out

thanx…

By: Steve

Steve — Thu, 20 Aug 2009 16:17:31 +0000

Varun—The convention followed by the Image Processing Toolbox is that gray-scale images represented as floating-point arrays have the range [0, 1], not [0, 255]. I recommend that you take a look at the intro sections in the Image Processing Toolbox users guide that describes the interactions between image type and data type conventions. Note also that the default display range can be overridden when you call imshow using this syntax:

imshow(A, [0 255])

Take a look at the documentation for imshow for more details.

By: Varun

Varun — Wed, 19 Aug 2009 20:51:40 +0000

Hey,
Steve..great post..I have just started using matlab and I stumbled over a problem..i.e
If i have an array (NxN..say 100X100) and i store random pixel values into it (in the range of 0-255) and do ‘imshow()’ should it show the different grey colors..I ran the code but it just showed white for values >0 and black for values=0…Why is it so? and how to I reflect the various greyscale color values of the array onto the image window..

By: Steve

Steve — Fri, 24 Jul 2009 01:01:55 +0000

Marco—Your question is pretty vague. What’s stopping you from dividing it in a certain number of parts? Presumably you know how you want to divide it, so divide away.

By: Marco

Marco — Thu, 23 Jul 2009 10:17:29 +0000

This technique is very interesting….but what if the image is not oroginally segmented as this one (I mean I have a single entity instead of multiple entities) and you want to divide it in a certain number of parts, and then ask the average value??

By: Steve

Steve — Wed, 27 May 2009 18:21:27 +0000

Santosh—Use something like this code to compute the mean of a subimage:

subimage = original_image(r1:r2, c1:c2);
average_value = mean2(subimage);

By: Santosh Bansode

Santosh Bansode — Sat, 09 May 2009 06:42:36 +0000

Hi there,
I have made a grey chart from black to white reducing the greyness. I have done it through normal Excel using fill color option, I printed it then and then took a picture of it. Now i want to read average value of pixels in different areas of grey level, probably in the centre of each grey area as there might be a bit of noise where a grey level meets another grey level. Please let me know how to do it. Kind Regards.

By: Steve

Steve — Fri, 08 May 2009 15:35:01 +0000

Allen—Yes, what you describe is possible. bwlabel and regionprops are the key functions.

By: allen j

allen j — Mon, 27 Apr 2009 23:32:36 +0000

Steve-
My question lies along the same as shawkat. You will have to excuse me as I stumbled upon this while doing some research for my thesis project. I am working on finding drag coefficients for a number of different odd objects. The dilemna is that finding the frontal area of these objects can be very difficult. One of my professors said that maybe doing a pixel count would then be able to tell be the area. I am following all of what you are doing with the coins, I just want to know if it is at all possible to help me find say the area of one of those coins, or if not at least the amount of pixels that make up one of the coins and then I could try to calculate area from there. I am more familiar with Mathcad as it is emphasized in our program, however we have matlab at school and so Im willing to learn if this would work. Let me know what you think. Thank you for your time and any help.