How to get the inverted mask in python?

Question

I'm fairly new to python and am struggling with a problem where I already spend quite some time on.

I work on detecting the object in an image and return a clipped version of it (using the SAM model). It works fine with one exception: I cannot get rid of the black background in the jpg file.

def segment_object(jpeg_base64, detected_object, predictor):
    jpeg_data = base64.b64decode(jpeg_base64)
    nparr = np.frombuffer(jpeg_data, np.uint8)
    image = cv2.imdecode(nparr, cv2.IMREAD_COLOR)

    predictor.set_image(image)

    bounding_poly = detected_object["bounding_poly"]["normalizedVertices"]
    image_width, image_height = image.shape[1], image.shape[0]
    vertices = [
        (int(vertex["x"] * image_width), int(vertex["y"] * image_height))
        for vertex in bounding_poly
    ]

    input_box = np.array(
        [
            min(vertices, key=lambda t: t[0])[0],
            min(vertices, key=lambda t: t[1])[1],
            max(vertices, key=lambda t: t[0])[0],
            max(vertices, key=lambda t: t[1])[1],
        ]
    )

    masks, _, _ = predictor.predict(
        point_coords=None,
        point_labels=None,
        box=input_box[None, :],
        multimask_output=False,
    )

    # Create a new image with a white background
    white_background = np.zeros_like(image, dtype=np.uint8)
    white_background[:] = (255, 255, 255)

    # Apply the mask returned by the predictor directly
    masked_image = cv2.bitwise_and(image, image, mask=masks[0].astype(np.uint8)) **works fine**

    
    # Combine the masked image and the white background
    not_mask = cv2.bitwise_not(masks[0].astype(np.uint8)) **issue**
    masked_white_background = cv2.bitwise_and(white_background, white_background, mask=not_mask)
    clipped_image = cv2.add(masked_image, masked_white_background)

    # Convert the OpenCV image to a PIL image and save it as a JPEG in a BytesIO object
    pil_image = Image.fromarray(cv2.cvtColor(clipped_image, cv2.COLOR_BGR2RGB))
    jpeg_buffer = io.BytesIO()
    pil_image.save(jpeg_buffer, format='JPEG')
    jpeg_bytes = jpeg_buffer.getvalue()

    # Encode the JPEG bytes as base64
    clipped_jpeg_base64 = base64.b64encode(jpeg_bytes).decode()

    return {
        'original_image': image_to_base64(image),
        'mask': image_to_base64(masks[0].astype(np.uint8)),
        'masked_image': image_to_base64(masked_image),
        'masked_white_background': image_to_base64(masked_white_background),
        'clipped_image': clipped_jpeg_base64
    }

More specifically: The prediction works, the masked_image is returned (but with black background). My idea is to use the inverted mask to make the rest white. However it won't work, the masked_white_background and clipped_image both return a completely white jpg file.

Has anyone an idea what the reason might be?

Answer 1

The issue seems to be this line

masked_white_background = cv2.bitwise_and(white_background, white_background, mask=not_mask)

OpenCV docs says that bitwise_and returns the conjunction of two input arrays where the keyword argument mask evaluates to true. So what you're essentially getting is the bitwise and of two white backgrounds.