Now, if we want to shift an image 7 pixels to the left and 23 pixels up, our translation matrix would look like the following: M = np.float32([Īnd as a final example, let’s suppose we want to translate our image 30 pixels to the left and 12 pixels down: M = np.float32([Īs you can see, defining our affine transformation matrix for image translation is quite easy!Īnd once our transformation matrix is defined, we can simply perform the image translation using the cv2.warpAffine function, like so: shifted = cv2.warpAffine(image, M, (image.shape, image.shape)) Our translation matrix would look like the following (implemented as a NumPy array): M = np.float32([ Positive values for will shift the image downįor example, let’s suppose we want to shift an image 25 pixels to the right and 50 pixels down.Negative values for shifts the image up.

Positive values for shifts the image to the right.

Negative values for the value will shift the image to the left.

Figure 1: To translate an image with OpenCV, we must first construct an affine transformation matrix.įor the purposes of translation, all we care about are the and values:

Photo translate