scannet_wrangling_scripts

Downloading and Extracting ScanNetv2

Developed and tested with python 3.9.

The included license at LICENSE applies only to reader.py and SensorData.py.

These scripts should help you export ScanNetv2 to the following format:

SCANNET_ROOT
    scans_test (test scans)
        scene0707
            scene0707_00_vh_clean_2.ply (gt mesh)
            sensor_data
                frame-000261.pose.txt
                frame-000261.color.jpg 
                frame-000261.color.512.png (optional, image at 512x384)
                frame-000261.color.640.png (optional, image at 640x480)
                frame-000261.depth.png (full res depth, stored scale *1000)
                frame-000261.depth.256.png (optional, depth at 256x192 also
                                            scaled)
            scene0707.txt (scan metadata and image sizes)
            intrinsic
                intrinsic_depth.txt
                intrinsic_color.txt

        ...
    scans (val and train scans)
        scene0000_00
            (see above)
        scene0000_01
        ....

Make sure all the packages in env.yml are installed in your environment.

Downloading ScanNetv2

The download_scannet.py script is from https://kaldir.vc.in.tum.de/scannet/download-scannet.py

Please make sure you fill in this form before downloading the data: https://kaldir.vc.in.tum.de/scannet/ScanNet_TOS.pdf

Download the dataset by running:

python download_scannet.py -o SCANNET_ROOT

For one scan debug use:

python download_scannet.py -o SCANNET_ROOT --id scene0707_00

This will download a .sens file, .txt file, the high resolution mesh ply, and a lower resolution mesh ply.

.txt will include meta information for the scan. See the next section for extracting the .sens file.

Extracting data from .sens files

Please use the intrinsics directly from the downloaded .txt file from the dataset.

This is a modified version of the SensReader python script at https://github.com/ScanNet/ScanNet/tree/master/SensReader/python

reader.py will extract depth, jpg, and intrinics files from ScanNetv2's downloaded .sens files. It will dump the jpg data directly to disk without uncompressing/compressing.

To extract all scans for test:

python reader.py --scans_folder SCANNET_ROOT/scans_test \
                 --output_path  OUTPUT_PATH/scans_test \
                 --scan_list_file splits/scannetv2_test.txt \
                 --num_workers 12 \
                 --export_poses \
                 --export_depth_images \
                 --export_color_images \
                 --export_intrinsics;

For train and val

python reader.py --scans_folder SCANNET_ROOT/scans \
                 --output_path  OUTPUT_PATH/scans \
                 --scan_list_file splits/scannetv2_train.txt \
                 --num_workers 12 \
                 --export_poses \
                 --export_depth_images \
                 --export_color_images \
                 --export_intrinsics;

python reader.py --scans_folder SCANNET_ROOT/scans \
                 --output_path  OUTPUT_PATH/scans \
                 --scan_list_file splits/scannetv2_val.txt \
                 --num_workers 12 \
                 --export_poses \
                 --export_depth_images \
                 --export_color_images \
                 --export_intrinsics;

OUTPUT_PATH can be the same directory as the ScanNet root directory SCANNET_ROOT.

For one scan use --single_debug_scan_id.

For caching resized pngs for depth and color files, run:

python reader.py --scans_folder SCANNET_ROOT/scans \
                 --output_path OUTPUT_PATH/scans \
                 --scan_list_file splits/scannetv2_train.txt \
                 --num_workers 12 \
                 --export_depth_images \
                 --export_color_images \
                 --rgb_resize 512 384 \
                 --depth_resize 256 192;

and for images at 640x480:

python reader.py --scans_folder SCANNET_ROOT/scans \
                 --output_path OUTPUT_PATH/scans \
                 --scan_list_file splits/scannetv2_train.txt \
                 --num_workers 12 \
                 --export_color_images \
                 --rgb_resize 640 480 \