My questions focus on the function of boxes3d_to_corners3d from lib\utils\kitti_utils as follow:
boxes_num = boxes3d.shape[0]
h, w, l = boxes3d[:, 3], boxes3d[:, 4], boxes3d[:, 5]
x_corners = np.array([l / 2., l / 2., -l / 2., -l / 2., l / 2., l / 2., -l / 2., -l / 2.], dtype=np.float32).T # (N, 8)
z_corners = np.array([w / 2., -w / 2., -w / 2., w / 2., w / 2., -w / 2., -w / 2., w / 2.], dtype=np.float32).T # (N, 8)
y_corners = np.zeros((boxes_num, 8), dtype=np.float32)
y_corners[:, 4:8] = -h.reshape(boxes_num, 1).repeat(4, axis=1) # (N, 8)
If these coordinates of corners add their according center coordinates, so the bounding box of object is equivalent to enhance abount h/2. Why are we doing this? The final object's position is wrong in my opinion.
Can someone help me?
My questions focus on the function of boxes3d_to_corners3d from lib\utils\kitti_utils as follow: boxes_num = boxes3d.shape[0] h, w, l = boxes3d[:, 3], boxes3d[:, 4], boxes3d[:, 5] x_corners = np.array([l / 2., l / 2., -l / 2., -l / 2., l / 2., l / 2., -l / 2., -l / 2.], dtype=np.float32).T # (N, 8) z_corners = np.array([w / 2., -w / 2., -w / 2., w / 2., w / 2., -w / 2., -w / 2., w / 2.], dtype=np.float32).T # (N, 8) y_corners = np.zeros((boxes_num, 8), dtype=np.float32) y_corners[:, 4:8] = -h.reshape(boxes_num, 1).repeat(4, axis=1) # (N, 8)
If these coordinates of corners add their according center coordinates, so the bounding box of object is equivalent to enhance abount h/2. Why are we doing this? The final object's position is wrong in my opinion. Can someone help me?