Skip to content

epalzeolithe/ABViewMain

BranchesTags

Repository files navigation

ABView - Aerobatic Flight Viewer

DEMO https://youtu.be/CWQwiJPiQm4

ABView is a desktop application for visualizing and analyzing aerobatic flight data. It synchronizes multiple video feeds with inertial (IMU), GPS, and accelerometer data recorded during aerobatic flights, providing a real-time multi-panel view of the aircraft's attitude, trajectory, and flight parameters.

Current version: 1.9 GPX import

Features

  • Dual Video Playback -- Synchronized side-by-side display of two onboard camera feeds (e.g. cockpit + external) with timestamp overlay.
  • 3D GPS Trajectory -- Interactive Matplotlib 3D plot showing the flight path color-coded by altitude, with ground projection, north indicator, and aerobatic box axis overlay.
  • 3D Attitude Visualization -- Real-time pygfx 3D scene rendering a CAP10 aircraft STL model oriented by quaternion data from the Insta360 X4 IMU.
  • Artificial Horizon HUD -- Two artificial-horizon widgets overlaid on the 3D view: a standard front-facing horizon and a wingtip (side-view) horizon for aerobatic reference.
  • Flight Parameters Overlay -- Live readouts of pitch, bank angle, GPS heading, groundspeed (color-coded by airspeed range), altitude, vertical speed (ft/min), and G-load (with min/max tracking).
  • OpenStreetMap Tracking -- Embedded Leaflet map showing the aircraft position in real time, with the aerobatic box axis drawn on the map.
  • Data Merge Pipeline -- Companion script (MERGE 0.9.py) that fuses data from an Insta360 X4 camera (gyroscope + accelerometer via Gyroflow / gyro2bb), a GNS3000 GPS receiver (NMEA), and an iPhone running SensorLog into a single time-aligned CSV.

Architecture Overview

The project consists of two main scripts:

File Role
MERGE 0.9.py Data preparation -- Extracts, converts, filters, and merges sensor data from multiple sources into a unified data/merged.csv file.
ABView 0.9.py Viewer application -- PyQt5 desktop GUI that loads the merged CSV alongside two MP4 video files and presents synchronized multi-panel visualization.

Data Flow

Insta360 X4 (.insv)
  |
  +---> Gyroflow CLI ---> quaternion CSV (30 fps)
  |                             |
  +---> gyro2bb tool ---> gyro + accel CSV (1000 Hz)
  |                             |
  +---- interpolate & merge ----+
                |
                v
         X4 DataFrame (100 Hz, decimated to 10 Hz)

GNS3000 GPS (.TXT, NMEA)
  |
  +---> parse GNGGA/GNRMC sentences ---> GPS DataFrame (4 Hz)

iPhone SensorLog (.csv)
  |
  +---> parse & decimate (100 Hz -> 20 Hz) ---> iPhone DataFrame

         |           |            |
         +--- merge_asof (timestamp) ---+
                      |
                      v
              data/merged.csv
                      |
                      v
               ABView 0.9.py  +  video1.mp4  +  video2.mp4

Requirements

System

  • Python 3.10+
  • macOS (primary target) or Linux with display server
  • Gyroflow binary installed (for data export from .insv files)

Python Dependencies

PyQt5
PyQtWebEngine
opencv-python
numpy
pandas
scipy
matplotlib
pygfx
pymediainfo

Install with:

pip install PyQt5 PyQtWebEngine opencv-python numpy pandas scipy matplotlib pygfx pymediainfo

External Tools

Tool Purpose Location
Gyroflow CLI Export quaternion metadata from Insta360 .insv files Configured in MERGE 0.9.py as GYROFLOW_BIN (default: /Applications/Gyroflow.app/Contents/MacOS/gyroflow)
gyro2bb Extract raw gyroscope and accelerometer data from .insv files Bundled in tool/gyro2bb-mac-arm64

Data Preparation

Input Files

Place the following files in the data/ directory:

File Source Description
VID_*.insv (x2) Insta360 X4 Raw video files containing embedded IMU data
LOG*.TXT GNS3000 GPS NMEA log file (GNGGA + GNRMC sentences)
sensorlog.csv iPhone SensorLog app Accelerometer, gyroscope, quaternion, and GPS data
CAP10.stl 3D model STL file of the aircraft for 3D visualization (already included)

Running the Merge

python "MERGE 0.9.py"

This will produce data/merged.csv containing time-aligned columns from all sensors.

Configuration (top of MERGE 0.9.py)

X4_INSV_1 = "data/VID_20260221_091717_00_050.insv"   # First .insv file
X4_INSV_2 = "data/VID_20260221_091717_00_051.insv"   # Second .insv file
GPS_GNS3000 = "data/LOG00003.TXT"                     # GNS3000 NMEA log
IPHONE_SENSORLOG = "data/sensorlog.csv"                # iPhone SensorLog export
OUTPUT = "data/merged.csv"                             # Output merged file

SKIP_X4_EXPORT = True       # Set to False on first run to export from .insv
SKIP_GNS3000_IMPORT = True  # Set to False on first run to parse NMEA
SKIP_IPHONE_IMPORT = True   # Set to False on first run to parse SensorLog

First run: Set all SKIP_* flags to False so the raw data is exported and parsed. On subsequent runs, set them to True to reuse cached intermediate files and speed up the process.

Running ABView

Input Files

Ensure the following files are in the data/ directory:

  • merged.csv -- Output from the merge step
  • video1.mp4 -- First camera video (e.g. cockpit)
  • video2.mp4 -- Second camera video (e.g. external / wing)
  • CAP10.STL -- Aircraft 3D model (included)

Launch

python "ABView 0.9.py"

Configuration (top of ABView 0.9.py)

DATA = "data/merged.csv"        # Merged sensor data
VIDEO1 = "data/video1.mp4"      # First video feed
VIDEO2 = "data/video2.mp4"      # Second video feed
STL_FILE = "data/CAP10.STL"     # 3D aircraft model

BOX = 0.007 * 1.5              # Map box size in degrees latitude
DF_FREQ = 100                   # DataFrame frequency (Hz)
TRACE = 9000                    # Trajectory trail length (samples)
VITESSE_MISE_EN_LIGNE = 80     # Line-up speed threshold (km/h)
PITCH_MONTAGE_PAR_DEFAUT = 15  # Default camera mounting pitch angle (degrees)

User Interface

Layout

+---------------------+---------------------+
|                     |                     |
|     Video 1         |     Video 2         |
|    (cockpit)        |    (external)       |
|                     |                     |
+----------+----------+----------+----------+
|                     |          |          |
|      3D View        | GPS 3D   |  OSM    |
|      (pygfx)        | Plot     |  Map    |
|                     |          |          |
+----------+----------+----------+----------+
|  [===================slider===========]   |
|  [Pause][Recalibrer][Start][Mise en       |
|   ligne][BOX][Zoom-][Zoom+][Reset]...     |
+-------------------------------------------+

3D View Overlays

The pygfx 3D panel displays the CAP10 model with the following HUD overlays:

Position Information Color
Top-right GPS Heading Black
Right (upper) Pitch angle Green
Right (middle) Bank angle Blue
Right (center) G-load (with min/max) Red (color varies with G)
Bottom-right Groundspeed Color-coded by speed range
Bottom-right Altitude (ft) Blue
Bottom-right Vertical speed (ft/min) Blue
Top-left Front artificial horizon --
Left Wingtip artificial horizon --

Keyboard Shortcuts

Key Action
Space Play / Pause
Home Go to start
Ctrl+Q Quit
Ctrl+M Go to line-up point (speed > 80 km/h)
Ctrl+B Go to BOX entry (altitude > 3000 ft)
Ctrl++ Zoom in (3D GPS plot)
Ctrl+- Zoom out (3D GPS plot)
Ctrl+0 Reset zoom
Ctrl+[ Decrease trajectory trail length
Ctrl+] Increase trajectory trail length
Ctrl+T Reset trajectory trail length

Mouse

  • Double-click on the 3D GPS plot to open it in full-screen (playback pauses automatically; resumes on close).
  • Click and drag on the 3D GPS plot to rotate the view (elevation is locked by default).

Project Structure

ABView/
  ABView 0.9.py       # Main viewer application (PyQt5)
  MERGE 0.9.py         # Data merge pipeline
  data/
    CAP10.stl          # 3D aircraft model (CAP10)
    merged.csv         # (generated) Merged sensor data
    video1.mp4         # (user-provided) First video
    video2.mp4         # (user-provided) Second video
    *.insv             # (user-provided) Insta360 X4 raw files
    LOG*.TXT           # (user-provided) GNS3000 GPS log
    sensorlog.csv      # (user-provided) iPhone SensorLog export
  tool/
    gyro2bb-mac-arm64  # gyro2bb binary for macOS ARM64
  docs/
    architecture.md    # Detailed architecture documentation
    data-pipeline.md   # Data pipeline documentation
    user-guide.md      # User guide

Version History

Version Notes
0.8 Working version, requires export via Gyroflow GUI
0.9 Direct export from X4 .insv files -- requires Gyroflow binary installed

License

This project is currently provided without a license. All rights reserved by the author.

Contributing

Contributions are welcome. Please open an issue or pull request on GitHub.

About

No description, website, or topics provided.

Resources

Readme
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages