Access Data Set Using ROS Python

Access Dataset using ROS/Python

Download

Download the dataset from here.

Dataset Structure

The dataset has the following folder structure.

MOD_Dataset
├── Text_Images 
|   ├── train_seq_room1_obj1
|   |   ├── masks.zip
|   |   ├── images.zip
|   |   ├── depthmaps.zip
|   |   ├── images.txt
|   |   ├── depthmaps.txt
|   |   ├── events.txt
|   |   ├── calib.txt
|   |   ├── camera_pose.txt
|   |   ├── obj_pose.zip
|   |   ├── obj_masks.zip
|   |   └── depthmaps.zip
|   ├── train_seq_room1_obj2
|   |   └── Same structure as train_seq_room1_obj1
|   ├── train_seq_room1_obj3
|   |   └── Same structure as train_seq_room1_obj1
|   ├── train_seq_room3_obj1
|   |   └── Same structure as train_seq_room1_obj1
|   ├── train_seq_room3_obj2
|   |   └── Same structure as train_seq_room1_obj1
|   ├── train_seq_room3_obj3
|   |   └── Same structure as train_seq_room1_obj1
|   └── test_seq
|       └── Same structure as train_seq_room1_obj1
└── Bag_Files
    ├── train_seq_room1_obj1.bag
    ├── train_seq_room1_obj2.bag
    ├── train_seq_room1_obj3.bag
    ├── train_seq_room3_obj1.bag
    ├── train_seq_room3_obj2.bag
    ├── train_seq_room3_obj3.bag
    └── test_seq.bag

• The images are in .png format and images.txt is a tab separated file with each row containing time \t ImageName information.
• depthmaps.txt has the same format as images.txt.
• events.txt is a space separated file with each row containiny time loc_x loc_y polarity.
• calib.txt contains the camera calibration data as fx fy cx cy k1 k2 r1 r2 (focal length in x, focal length in y, principle point in x, principle point in y, four rad-tan distrotion coefficients).
• The Depthmaps are in .exr format and can be loaded using the OpenEXR package.
• Note - Object masks, Object poses and Camera poses are only available in Text_Image folder.