语义分割3D

标注文件目录结构

+--- 1599625710056
|   +--- 1599625710056.pcd
|   +--- 1599625710056.json
+--- 1599625740054
|   +--- 1599625740054.pcd
|   +--- 1599625740054.json

推理文件目录结构

+--- 1599625710056
|   +--- 1599625710056.json
+--- 1599625740054
|   +--- 1599625740054.json

3D语义分割-Octopus.json

{
  "labels": [
    {
      "label_meta_id": 1,
      "polygon_3d_v2": {
        "ascii_char": "\""
      },
      "shape_type": "polygon_3d_v2",
      "serial_number": 0,
      "label_object_id": -1,
      "name": "car"
    },
    {
      "label_meta_id": 19,
      "polygon_3d_v2": {
        "ascii_char": "4"
      },
      "shape_type": "polygon_3d_v2",
      "serial_number": 0,
      "label_object_id": -1,
      "name": "traffic-sign"
    }
  ],
  "labels_ext": {
    "ascii_string": "000000000000000000000000000000000000000000000000000000000000000000444404444000111000000000000000000000000000000000000000000000000,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,"
  }
}

标注文件目录结构

+--- labels
|   +--- 000000.label
|   +--- 000001.label
+--- velodyne
|   +--- 000000.bin
|   +--- 000001.bin

推理文件目录结构

+--- 000000.label
+--- 000001.label

3D语义分割-semanticKITTI.label

>>> pre = np.fromfile("000000.label", np.uint32) 
>>> pre 
array([ 0,  0,  0, ..., 48, 48, 48], dtype=uint32) 
>>> np.unique(pre) 
array([       0,        1,       30,       40,       44,       48, 
              50,       51,       52,       70,       71,       72,
              80,       81,       99,    65788,    65790,   131324,
          131326,   196860,   196862,   589854,   983070,  1572875,
         1638411, 12713994, 12779530, 12845066, 12976138], dtype=uint32) 
>>> post = pre & 0xFFFF 
>>> post 
array([ 0,  0,  0, ..., 48, 48, 48], dtype=uint32) 
>>> np.unique(post) 
array([  0,   1,  10,  11,  30,  40,  44,  48,  50,  51,  52,  70,  71,
        72,  80,  81,  99, 252, 254], dtype=uint32)