How to resize the image canvas to maintain a square aspect ratio

Question

I'd like to get a 1000 x 1000 picture in Python from any input picture so that the input doesn't lose its aspect ratio. In other words, I want to resize the input so that its longer dimension be 1000 pixels and "fill" the other dimension with the background color until it becomes 1000 x 1000 square. The original must be in the center at the end.

Answer 1

down voteacceptedUsing OpenCVYou can use resize() in OpenCV to resize the image up/down to the size you need. However, resize() requires that you put in either the destination size (in both dimensions) or the scaling (in both dimensions), so you can't just put one or the other in for 1000 and let it calculate the other for you. So the most robust way to do this is to find the aspect ratio and calculate what the smaller dimension would be when the bigger one is stretched to 1000. Then you can resize.

down voteaccepted

Answer 2

Building on Alexander-Reynolds answer above, here is the code that handles all possible sizes and situations.

def resizeAndPad(img, size, padColor=255):

h, w = img.shape[:2]
sh, sw = size

# interpolation method
if h > sh or w > sw: # shrinking image
    interp = cv2.INTER_AREA

else: # stretching image
    interp = cv2.INTER_CUBIC

# aspect ratio of image
aspect = float(w)/h 
saspect = float(sw)/sh

if (saspect > aspect) or ((saspect == 1) and (aspect <= 1)):  # new horizontal image
    new_h = sh
    new_w = np.round(new_h * aspect).astype(int)
    pad_horz = float(sw - new_w) / 2
    pad_left, pad_right = np.floor(pad_horz).astype(int), np.ceil(pad_horz).astype(int)
    pad_top, pad_bot = 0, 0

elif (saspect < aspect) or ((saspect == 1) and (aspect >= 1)):  # new vertical image
    new_w = sw
    new_h = np.round(float(new_w) / aspect).astype(int)
    pad_vert = float(sh - new_h) / 2
    pad_top, pad_bot = np.floor(pad_vert).astype(int), np.ceil(pad_vert).astype(int)
    pad_left, pad_right = 0, 0

# set pad color
if len(img.shape) is 3 and not isinstance(padColor, (list, tuple, np.ndarray)): # color image but only one color provided
    padColor = [padColor]*3

# scale and pad
scaled_img = cv2.resize(img, (new_w, new_h), interpolation=interp)
scaled_img = cv2.copyMakeBorder(scaled_img, pad_top, pad_bot, pad_left, pad_right, borderType=cv2.BORDER_CONSTANT, value=padColor)

return scaled_img