with numpy version >=1.20 the np.float is changed to np.float64 which cause error while using the cython_bbox package. So i have fixed it.
This is already fixed in 17