[data_aug.py] 数据增强时，当检测当前bbox与从database采样的bbox是否有碰撞时，代码疑似有误？

[eggman-1024]

3. box_collision_test

    avoid_coll_boxes_bv_corners = bbox3d2bevcorners(avoid_coll_boxes)
    sampled_cls_bboxes_bv_corners = bbox3d2bevcorners(sampled_cls_bboxes)
    coll_query_matrix = np.concatenate([avoid_coll_boxes_bv_corners, sampled_cls_bboxes_bv_corners], axis=0)
    coll_mat = box_collision_test(coll_query_matrix, coll_query_matrix)

 这里的box_collision_test()接收的参数为什么两个是相等的？