Python API 2.0 MImage

[Anonymous]

In the API 2.0 Mimage.pixels() returns a long pointer to the pixel data. How do I get to the actual pixel data now that 2.0 does not use MScriptUtil? I am brand new to using the API and its a bit confusing.

[cheng_xi_li]

Hi leo,

You can use uuid.ctypes.cast to wrap it as a c_char and then use uuid.ctypes.string_at to create a cstring for it. The size of the string should be width * height * depth(4).

...
import uuid

...
longPointer = image.pixels()
pixel = uuid.ctypes.cast(longPointer, uuid.ctypes.POINTER(uuid.ctypes.c_char))
pixels_size = view.portWidth() * view.portHeight() * 4 # current depth in the document.
pixels_string = uuid.ctypes.string_at(pixel, pixels_size)

For more details, please look at the ctypes section in the Python official document.

Yours,

Li

[Anonymous]

I had the same problem - thanks for the answer!

How would I go about actually getting to the values of those pixels?

I understand that one can create a QImage from the pixels_string - but can I turn that directly into a python array as well?

Thanks!

seb

[Anonymous]

I ended up just doing it with only Qt like this

def image_to_array(image, width, height):
    '''
    return array of integer 0-255 rgba values
    [(r, g, b, a)]
    '''
    img = QtGui.QImage(width, height, QtGui.QImage.Format.Format_ARGB32)
    img.load(image)
    colorArray = []
    for y in range(height):
        for x in range(width):
            color = QtGui.QColor()
            color.setRgba(img.pixel(x,y))
            colorArray.append(color.getRgb())
    return colorArray

[Anonymous]

Makes sense 🙂 At least it's still maya only and doesn't require something like PIL!

S