Missing the camera to lidar extrinsic in 3dbox_to_cloud.py

[yaobaishen]

There are two steps transformation in the 3dbox_to_cloud.py:

1. object coordinate to camera coordinate
2. camera coordinate to lidar coordinate
   The bug exists in the second step, it doesn't consider the rotation and translation between camera coord and lidar coord axis, so may leads to mismatch in customized kitti data, just like below:
   The issue could disappear if we add the camera to lidar transformation matrix back.