I am used to Java's implementation of OpenCV. I want to create a Mat structure, fill data into it, extract a submat and then apply some image transform. In Java, I use:
my_mat = new Mat(my_rows, my_cols, CvType.CV_8U);
my_mat.put(0, 0, my_data);
my_mat.submat(0, my_other_rows, 0, my_other_cols);
But I didn't find anything working in python's OpenCV. I found this link on the OpenCV forums: cv2.CreateMat in python, but the link is broken
Even though the question was asked a long time ago, this should help anyone looking for an answer today.
This should be applicable to opencv versions >= 2.X. In python, the opencv images are now represented as numpy arrays. So, a Mat object can simply be created as follows:
cvImg = np.zeros((rows, columns, channels), dtype = "uint8")
For OpenCV 1.x :
You can use CreateMat to do that :
Creates a matrix header and allocates the matrix data.
Python: cv.CreateMat(rows, cols, type) → mat
Parameters:
rows – Number of rows in the matrix
cols – Number of columns in the matrix
type – The type of the matrix elements in the form CV_<bit depth><S|U|F>C<number of channels> , where S=signed, U=unsigned, F=float. For example, CV _ 8UC1 means the elements are 8-bit unsigned and the there is 1 channel, and CV _ 32SC2 means the elements are 32-bit signed and there are 2 channels.
The function call is equivalent to the following code:
CvMat* mat = cvCreateMatHeader(rows, cols, type);
cvCreateData(mat);
For cv2 interface :
The new cv2 interface for Python integrates numpy arrays into the OpenCV framework, which makes operations much simpler as they are represented with simple multidimensional arrays. here's a starting example :
import numpy as np, cv
vis = np.zeros((384, 836), np.float32)
h,w = vis.shape
vis2 = cv.CreateMat(h, w, cv.CV_32FC3)
vis0 = cv.fromarray(vis)
AttributeError: 'module' object has no attribute 'CreateMat'...OpenCV 3.0
