[enhance]: Use points loader to load point cloud data in db_infos (#87)

* Change gt_database and dbsampler.

* Add int condition for use_dim.

* Change default sweep_num.

* Change docstring.

* Change to pipeline in db_sampler.

* remove comments.

mmdet3d/datasets/pipelines/dbsampler.py

 import copy
+import mmcv
 import numpy as np
 import os
-import pickle
 
 from mmdet3d.core.bbox import box_np_ops
 from mmdet3d.datasets.pipelines import data_augment_utils
+from mmdet.datasets import PIPELINES
 from ..registry import OBJECTSAMPLERS
 
 
@@ -86,6 +87,8 @@ class DataBaseSampler(object):
         prepare (dict): Name of preparation functions and the input value.
         sample_groups (dict): Sampled classes and numbers.
         classes (list[str]): List of classes. Default: None.
+        points_loader(dict): Config of points loader. Default: dict(
+            type='LoadPointsFromFile', load_dim=4, use_dim=[0,1,2,3])
     """
 
     def __init__(self,
@@ -94,7 +97,11 @@ class DataBaseSampler(object):
                  rate,
                  prepare,
                  sample_groups,
-                 classes=None):
+                 classes=None,
+                 points_loader=dict(
+                     type='LoadPointsFromFile',
+                     load_dim=4,
+                     use_dim=[0, 1, 2, 3])):
         super().__init__()
         self.data_root = data_root
         self.info_path = info_path
@@ -103,9 +110,9 @@ class DataBaseSampler(object):
         self.classes = classes
         self.cat2label = {name: i for i, name in enumerate(classes)}
         self.label2cat = {i: name for i, name in enumerate(classes)}
+        self.points_loader = mmcv.build_from_cfg(points_loader, PIPELINES)
 
-        with open(info_path, 'rb') as f:
-            db_infos = pickle.load(f)
+        db_infos = mmcv.load(info_path)
 
         # filter database infos
         from mmdet3d.utils import get_root_logger
@@ -244,8 +251,9 @@ class DataBaseSampler(object):
                 file_path = os.path.join(
                     self.data_root,
                     info['path']) if self.data_root else info['path']
-                s_points = np.fromfile(
-                    file_path, dtype=np.float32).reshape([-1, 4])
+
+                results = dict(pts_filename=file_path)
+                s_points = self.points_loader(results)['points']
                 s_points[:, :3] += info['box3d_lidar'][:3]
 
                 count += 1

tools/data_converter/create_gt_database.py

@@ -145,7 +145,7 @@ def create_groundtruth_database(dataset_class_name,
         type=dataset_class_name,
         data_root=data_path,
         ann_file=info_path,
-    )
+        use_valid_flag=True)
     if dataset_class_name == 'KittiDataset':
         file_client_args = dict(backend='disk')
         dataset_cfg.update(
@@ -176,7 +176,9 @@ def create_groundtruth_database(dataset_class_name,
             dict(
                 type='LoadPointsFromMultiSweeps',
                 sweeps_num=10,
-            ),
+                use_dim=[0, 1, 2, 3, 4],
+                pad_empty_sweeps=True,
+                remove_close=True),
             dict(
                 type='LoadAnnotations3D',
                 with_bbox_3d=True,
@@ -191,7 +193,6 @@ def create_groundtruth_database(dataset_class_name,
                                      f'{info_prefix}_dbinfos_train.pkl')
     mmcv.mkdir_or_exist(database_save_path)
     all_db_infos = dict()
-
     if with_mask:
         coco = COCO(osp.join(data_path, mask_anno_path))
         imgIds = coco.getImgIds()