Detect Camera shift from images snapshots taking every hour with OpenCV

I have a couple of CCTV cameras taking snapshots every hour and saving them into a folder with the date and time. The scene (Field of view) of each camera doesn't change. I want to detect if the camera was shifted up/bottom or left/right from the previous view by comparing two snapshots.

this is similar to what I am trying to do: Python: Detect camera shift angle comparing the reference image and live image taken from a camera

But the solution is not clear, and the post is old.

How Can I apply the optical flow to compare two images and detect the shift of the scene without detecting the object's movement inside?

this is what I tried so far

    # optical flow between two frames
def optical_flow(old_frame, frame):
    old_gray = cv2.cvtColor(old_frame, cv2.COLOR_BGR2GRAY)
    frame_gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    # Create a mask image for drawing purposes
    mask = np.zeros_like(old_frame)
    mask_features = np.zeros_like(old_gray)
    mask_features[:,0:20] = 1
    mask_features[:,620:640] = 1

    # params for ShiTomasi corner detection
    # maxCorners is the number of pixels from the corner
    feature_params = dict( maxCorners = 800,
                       qualityLevel = 0.3,
                       minDistance = 3,
                       blockSize = 7,
                       mask = mask_features)

    # Parameters for lucas kanade optical flow
    lk_params = dict( winSize  = (15,15),
                  maxLevel = 2,
                  criteria = (cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, 10, 0.03))

     corners = init_new_features(old_gray, feature_params)

    while True:
        try:
            cam_moved = False
                    
            # calculate optical flow
            p1, st, err = cv2.calcOpticalFlowPyrLK(old_gray, frame_gray, corners, None, **lk_params)

                   
            good_new = p1[st==1]
            good_old = corners[st==1]
        
            # draw the tracks
            for i,(new,old) in enumerate(zip(good_new,good_old)):
                a,b = new.ravel()
                c,d = old.ravel()

                distance = calculateDistance(a,b,c,d)
                #print('Distance: ',int(distance))
            
        
                if distance > 5.0:
                    cam_moved = True
                    # update the previous frame and previous points
                    old_gray = frame_gray.copy()
                    corners = init_new_features(old_gray, feature_params)
                else:
                    old_gray = frame_gray.copy()
                    corners = good_new.reshape(-1,1,2)
                
                #mask = cv2.line(mask, (int(a),int(b)),(int(c),int(d)), color[i].tolist(), 2)
                #frame = cv2.circle(frame,(int(a),int(b)),5,color[i].tolist(),-1)


            return cam_moved
        except TypeError as e:
            print(e)
            break

Thank you for any advice or help you can provide.