OpenCV Individual Pixel Access

Question

I'm trying to learn OpenCV and have been going through various tutorials and books. I've ran into an odd problem. I have a 3 row, 3 column 3-channel matrix in OpenCV I named test. I filled it with dummy data like so:

Mat test(3,3, CV_8UC3, Scalar(50,100,200));

When I print out the matrix I get:

[[[50, 100, 200], [50, 100, 200], [50, 100, 200]], 
  [[50, 100, 200], [50, 100, 200], [50, 100, 200]], 
  [[50, 100, 200], [50, 100, 200], [50, 100, 200]]]

So far so good.

I want to access each individual RGB values for row 1, column 1. I read in the documentation and other answers on StackOverFlow to do the following for individual pixel access:
test.at(1,1)[0];

When I print out the values I get:

test.at<Vec3b>(1,1)[0] = 2
test.at<Vec3b>(1,1)[1] = d
test.at<Vec3b>(1,1)[2] = /310

Any ideas on what I'm doing wrong? I'm running OpenCV 2.10 on a Mac. Language is C++.

Thanks!

Answer 1

The values you are printing out are the ASCII characters for the integer values stored.

For example if you look up the ASCII table, say at this url, http://simple.wikipedia.org/wiki/ASCII

You will see that decimal 50 is equivalent to the character '2' and decimal 100 is 'd'.

In C++ you will get this result if you set a value into an int variable and then cast that int to a char.