I am trying to create a top-down view of an image. Currently I have some known source points and known destination points:
src_pts = np.array([(1223,1511),(2238,1090),(1300,1115),(1330,943),(360,1139),(1731,811),(982,831)])
dst_pts = np.array([(636,1982),(1045,1498),(637,1498),(637,1040),(226,1498),(1045,102),(225,102)])
I can transform the image using:
h, status = cv2.findHomography(src_pts, dst_pts)
img_out1 = cv2.warpPerspective(img, h, (img_dst.shape[1], img_dst.shape[0]))
Question:
Now I have coordinates of an additional source point where I dont know the coordinates in the destination image. Is it possible to compute the coordinats of the point in the destination image using the computed homography?