Running on previously collected data

To run the tracking.py program on previously recorded data, there are 2 options:

Pass the CLI arguments to the program with required details.
Mount configuration files for Radar and/or Video processing with required details.

Note

All options below, will assume you have pulled the docker image, and tagged it as tracking-image to simplify the following commands (so they're not platform dependant). Please see the docker quickstart section for more details.

Radar

Using CLI Arguments to Run on Collected Data

To run on collected data by changing CLI arguments, you can do that following.

Run Container with Data Mount To access data within the container you'll need to mount the local data into it. This example mounts the pwd/data folder into the /data location in the container. These can both be configured as you'd like.
LinuxJetson5
docker run -it tracking-image -v "$(pwd)"/data:/data
docker run --ipc=host --runtime=nvidia -it tracking-image -v "$(pwd)"/data:/data

Run the tracking program, specify the location to the data

python3 tracking.py --radar-from-file --radar-from-file <path-to-file>

Mounting Configuration Files

Mount the RadarConfig.yaml file to the /configuration section of the container when running. You may also want to mount /data to the container to easily pass data into it.

Running Container with Configuration/Data Mounts

LinuxJetson5

docker run -it tracking-image -v "$(pwd)"/configuration:/configuration -v "$(pwd)"/data:/data

docker run --ipc=host --runtime=nvidia -it tracking-image -v "$(pwd)"/configuration:/configuration -v "$(pwd)"/data:/data

Modify these 2 settings in the RadarConfig.yaml configuration file
```
# Run configuration
run: RERUN # Changed to RERUN
sourcePath: "/data/radar/run2-TD/" # Path to the data to be processed for RERUN mode
```
For full details on the RadarConfig.yaml file please see the configuration section.
Run the tracking program, specify the configuration file location. By default it's set to /configuration/RadarConfig.yaml.
```
python3 tracking.py --radar-config <path-to-config-file>
```

Video

CLI Arguments to Run on Collected Data

Run Container with Data Mount To access data within the container you'll need to mount the local data into it. This example mounts the pwd/data folder into the /data location in the container. These can both be configured as you'd like.
LinuxJetson5
docker run -it tracking-image -v "$(pwd)"/data:/data
docker run --ipc=host --runtime=nvidia -it tracking-image -v "$(pwd)"/data:/data
Run the tracking program, specify the location to the data
```
python3 tracking.py --video-source <path-to-file>
```

Mounting Configuration Files

Mount the VideoConfig.yaml file to the /configuration section of the container when running. You may also want to mount /data to the container to easily pass data into it.

Running Container with Configuration/Data Mounts

LinuxJetson5

docker run -it tracking-image -v "$(pwd)"/configuration:/configuration -v "$(pwd)"/data:/data

docker run --ipc=host --runtime=nvidia -it tracking-image -v "$(pwd)"/configuration:/configuration -v "$(pwd)"/data:/data

Modify these 2 settings in the VideoConfig.yaml configuration file
```
# Yolo Options
videoSource: "<absolutePathToFileOnDisk>"
```
For full details on the RadarConfig.yaml file please see the configuration section.
Running the tracking command. Optionally, if you mounted the configuration at a different path you can specify that path too. By default it's set to /configuration/VideoConfig.yaml.
```
python3 tracking.py --video-config <my-data-path>
```