How to use a NumPy slice as a function argument?-CodePudding

Currently extracting frames from a video to be processed later on. I have different areas of the video frames that I want to process so I need to be able to pass a numpy slice into a variable to be used as the function's argument. What's the syntax for that?

frame[5:49, 879:938] is the slice that I would like to use in a variable to be passed as the 3rd argument of the extractFrames function.

Here's my code:

def extractFrames(video_file, folder):
    try:
        os.mkdir(folder)
    except:
        shutil.rmtree(folder)
        os.mkdir(folder)
    capture = cv2.VideoCapture(video_file)
    count = 0
    while (capture.isOpened()):
        # Capture frame-by-frame
        ret, frame = capture.read()
        if ret == True:
            print('Read %d frame: ' % count, ret)
            cv2.imwrite(os.path.join(folder, "{:d}.jpg".format(count)), frame[5:49, 879:938])  # save frame as JPEG file
            count  = 1
        else:
            break
    # When everything done, release the capture
    capture.release()
    cv2.destroyAllWindows()

CodePudding user response：

When you say frame[5:49, 879:938] you are indexing into frame using a pair of slices. The slice type is built in to Python, and you can pass two of them to your function like this:

extractFrames("filename", "folder", x=slice(5, 49), y=slice(879, 938))

Defining your function like this:

def extractFrames(video_file, folder, x=slice(None), y=slice(None)):
    ...frame[x_slice, y_slice]...

The default arguments will take the whole frame...you can omit them if you want the slicing to be mandatory.

CodePudding user response：

You don't need to construct slice objects and pass those around.

You can just pass the bounds around:

def extractFrames(..., bounds):
    (y0, y1, x0, x1) = bounds
    ...
    subregion = frame[y0:y1, x0:x1]
    ....

extractFrames(..., (5, 49, 879, 938))

This also allows you to pass None for any bounds, so the slice becomes an "all-slice".

Or pass (x, y, w, h)-style numbers, if you have those anyway, and then slice [y:y h, x:x w].