Skip to content

mars_logger.md

Latest commit

 

History

History
175 lines (126 loc) · 5 KB

mars_logger.md

File metadata and controls

175 lines (126 loc) · 5 KB
Raw

MarsLogger — Android sensor capture for VIO/SLAM

Mars Logger records synchronized camera frames + accelerometer + gyroscope from an Android phone in a format suitable for visual-inertial odometry / SLAM benchmarks.

1. Install the app on the phone

  1. Grab the latest .apk from the project's transfer guide and install it.
  2. Enable Developer Options: Settings → About → tap Build number 7 times.
  3. Enable USB debugging under Developer Options.

2. Pull recordings from the phone

Install ADB on the host:

sudo apt-get update
sudo apt-get install -y android-tools-adb

Recordings are written to /sdcard/Android/data/edu.osu.pcv.marslogger/files/data. Browse first:

adb shell
cd /sdcard/Android/data/edu.osu.pcv.marslogger/files/data
ls
exit

Pull a session to the host:

MARS_DIR=/sdcard/Android/data/edu.osu.pcv.marslogger/files/data
OUTPUT_DIR=$HOME/Desktop/MarsLogger
mkdir -p "$OUTPUT_DIR" && cd "$OUTPUT_DIR"
adb pull "$MARS_DIR"

A session directory contains:

  • movie.mp4 — H.264 video
  • gyro_accel.csv — IMU samples (gyro + accel, hardware-timestamped)
  • frame_timestamps.txt — per-frame capture timestamps
  • intrinsics.yaml, *.csv metadata

3. Convert a recording to a ROS bag

Use kalibr_bagcreater.py from vio_common. It produces /cam0/image_raw and /imu0 topics with frame timestamps aligned to the IMU clock.

BAG_PYTHON=$HOME/workspace/vio_common/python/kalibr_bagcreater.py
ANDROID_DATA_DIR=$HOME/Desktop/MarsLogger/data/2023_03_07_10_16_29

python "$BAG_PYTHON" \
  --video           "$ANDROID_DATA_DIR/movie.mp4" \
  --imu             "$ANDROID_DATA_DIR/gyro_accel.csv" \
  --video_time_file "$ANDROID_DATA_DIR/frame_timestamps.txt" \
  --output_bag      "$ANDROID_DATA_DIR/movie.bag"

For high-resolution captures, allow taller frames:

python "$BAG_PYTHON" \
  --video                  "$ANDROID_DATA_DIR/movie.mp4" \
  --max_video_frame_height 10000 \
  --imu                    "$ANDROID_DATA_DIR/gyro_accel.csv" \
  --video_time_file        "$ANDROID_DATA_DIR/frame_timestamps.txt" \
  --output_bag             "$ANDROID_DATA_DIR/movie.bag"

4. Camera + IMU calibration with Kalibr

Kalibr is most easily run from its official Docker image.

Supported camera models:

pinhole-radtan, pinhole-equi, pinhole-fov,
omni-none,     omni-radtan,
eucm-none, ds-none

4.1 Launch the container

FOLDER=$HOME/Desktop/MarsLogger/data/2023_03_06_15_20_28

xhost +local:root
docker run -it \
  -e DISPLAY -e QT_X11_NO_MITSHM=1 \
  -v /tmp/.X11-unix:/tmp/.X11-unix:rw \
  -v "$FOLDER:/data" \
  kalibr

Inside the container:

source devel/setup.bash

4.2 Camera intrinsics

rosrun kalibr kalibr_calibrate_cameras \
  --bag    /data/cam_april.bag \
  --target /data/april_6x6.yaml \
  --models pinhole-radtan \
  --topics /cam0/image_raw

Output: cam_april-camchain.yaml.

4.3 Camera–IMU extrinsics

rosrun kalibr kalibr_calibrate_imu_camera \
  --bag    /data/cam_april.bag \
  --cam    /data/cam_april-camchain.yaml \
  --imu    /data/imu.yaml \
  --target /data/april_6x6.yaml

4.4 Common errors

[TargetViewTable] Tried to add second view to a given cameraId & timestamp. Maybe try to reduce the approximate syncing tolerance.

Two image messages share a timestamp (Kalibr de-dupes by (camera_id, timestamp)). Either re-record at a lower frame rate, or pass --dont-show-report --time-calibration --max-iter <N> plus a tighter sync tolerance.

5. ORB-SLAM3

cd $HOME/Desktop/MarsLogger/data/2023_03_07_10_16_29

# Mono
rosrun ORB_SLAM3 Mono \
  $HOME/workspace/ORB_SLAM3/Vocabulary/ORBvoc.txt \
  ./ORB3_EuRoC.yaml

# Mono + IMU
rosrun ORB_SLAM3 Mono_Inertial \
  $HOME/workspace/ORB_SLAM3/Vocabulary/ORBvoc.txt \
  ./ORB3_EuRoC.yaml

# Replay the bag with topics remapped to ORB-SLAM3's defaults
rosbag play movie.bag /cam0/image_raw:=/camera/image_raw /imu0:=/imu

The YAML must contain the camera intrinsics + camera-IMU extrinsics produced by Kalibr in §4.

6. VINS-Fusion

roslaunch vins vins_rviz.launch
rosrun vins vins_node $HOME/Desktop/MarsLogger/data/2023_03_06_15_20_28/android_imu_config.yaml

# In another terminal
rosbag play movie.bag

The VINS config YAML expects body_T_cam0, accelerometer/gyroscope noise densities, and the topic names that match the bag.

References