Extract GoPro highlights and GPMF data.

Overview

About gopro2gpx

Python script that parses the gpmd stream for GOPRO moov track (MP4) and extract the GPS info into a GPX (and kml) file.

Tested on a GoPro7, but it should work on any camera above the GoPro5.

Tested on Windows7 and MacOS Sierra.

I always like to print some additional info overlaying my action videos. These overlays will show data about speed, elevation, gps info, and so on. I started a project wrotten on python 2.7 that works fine. gets a gpx track file, a mp4 file and some configuration, and builds an overlay with the data:

Recently, I bought a Gopro7 black, that supports GPS inside the camera. Gopro's Quik allows you to process the metadata stored along the videos, and show it as an overlay. I want to extract the data from the MP4 file, so I start to read about in the web. The data is stored in a format called *GPMF, you can get all the info in this repo https://github.com/gopro/gpmf-parser. Also, there are some implementations in go. Check this repo https://github.com/stilldavid/gopro-utils/.

My idea is process the file in python, extract the data, and build a file in a known format. I generate kml and gpx.

Dependencies

Installation

  1. Package installation: there are two ways to install the package: a) Install via pip (with git installed):
pip install git+https://github.com/juanmcasillas/gopro2gpx

b) Or Download the repository, unpack it and instal with

python setup.py install
  1. Ensure you have FFmpeg and FFprobe installed in your system.

  2. If ffmpeg is not installed in a PATH location, the path can be specified in the config file.

    The configuration file is located in:

    • Windows: %APPDATA%\gopro2gpx\gopro2gpx.conf
    • Unix (Linux, Mac): $HOME/.config/gopro2gpx.conf ($XDG_CONFIG_HOME/gopro2gpx.conf to be exact)

    The configuration file has to look like this:

[ffmpeg]
ffmpeg = /path/to/ffmpeg
ffprobe = /path/to/ffprobe
  1. The script can then be invoced with
gopro2gpx

or

python3 -m gopro2gpx

(exchange python3 with your specific python installation)

E.g. to run it on the example data (skip bad points, show the labels debug, create hero6.kml and hero6.gpx files):

gopro2gpx -s -vvv samples/hero6.mp4 hero6
  1. With custom path for FFMPEG
export PATH=$PATH:/usr/local/opt/ffmpeg/bin
gopro2gpx -vvv samples/8/GH010159.MP4 output.bin

Arguments and options

% python gopro2gpx.py  --help
usage: gopro2gpx.py [-h] [-v] [-b] [-s] file outputfile

positional arguments:
  file           Video file or binary metadata dump
  outputfile     output file. builds KML and GPX

optional arguments:
  -h, --help     show this help message and exit
  -v, --verbose  increase output verbosity
  -b, --binary   read data from bin file
  -s, --skip     Skip bad points (GPSFIX=0)
  • file: Gopro MP4 file or binary file with the gpmd dump.
  • outputfile: Dump the GPS info into outputfile.kml and outputfile.gpx. Don't use extension.
  • -v, -vv, -vvv: Verbose mode. First show some info, second dumps the gpmd track info a file called outputfile.bin and third (-vvv) shows the labels.
  • -b: read the data from a binary dump fo the gpmd track istead of the MP4 video. Useful for testing, so I don't need to move big MP4 files.
  • -s: skip "bad" GPS points. When GPSFIX=0 (no GPS satellite signal received) GPS data is unacurrate. Ignore these points.

How to get "accurate" GPS data in your GoPro [IMPORTANT]

Follow these steps:

  1. Turn the camera on using the side mode button (Do Not use Quick Capture if GPS information is important to you)
  2. Wait at least one minute (5-10 is best) before you start to record to allow the most satellite locks
  3. Use the camera in the Frame mount and not the Super Suit (it will work in the Super Suit but might have a weaker signal)

Read the following thread for more info:

Technical info

To get the gpmd data, we need to explore the MP4 container, and extract the stream marked as gpmd. The script does it automatically, but here is the output from ffprobe:

    % ffprobe GH010039.MP4

    The channel marked as gpmd (Stream #0:3(eng): Data: none (gpmd / 0x646D7067), 29 kb/s (default))
    In this case, the stream #0:3 is the required one (get the 3)

    Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'GH010039.MP4':
    Metadata:
        major_brand     : mp41
        minor_version   : 538120216
        compatible_brands: mp41
        creation_time   : 2019-02-10 10:59:19
    Duration: 00:00:21.80, start: 0.000000, bitrate: 60420 kb/s
        Stream #0:0(eng): Video: h264 (High) (avc1 / 0x31637661), yuvj420p(pc, bt709), 2704x1520 [SAR 1:1 DAR 169:95], 
        60173 kb/s, 50 fps, 50 tbr, 90k tbn, 100 tbc (default)
        Metadata:
        creation_time   : 2019-02-10 10:59:19
        handler_name    : GoPro AVC  
        encoder         : GoPro AVC encoder
        timecode        : 10:59:19:31
        Stream #0:1(eng): Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz, stereo, fltp, 189 kb/s (default)
        Metadata:
        creation_time   : 2019-02-10 10:59:19
        handler_name    : GoPro AAC  
        timecode        : 10:59:19:31
        Stream #0:2(eng): Data: none (tmcd / 0x64636D74), 0 kb/s (default)
        Metadata:
        creation_time   : 2019-02-10 10:59:19
        handler_name    : GoPro TCD  
        timecode        : 10:59:19:31
        Stream #0:3(eng): Data: none (gpmd / 0x646D7067), 29 kb/s (default)
        Metadata:
        creation_time   : 2019-02-10 10:59:19
        handler_name    : GoPro MET  
        Stream #0:4(eng): Data: none (fdsc / 0x63736466), 12 kb/s (default)
        Metadata:
        creation_time   : 2019-02-10 10:59:19
        handler_name    : GoPro SOS  

We need the stream called in this clase, #0:3(eng) that is, the 0:3 stream:

        [...]
        Stream #0:3(eng): Data: none (gpmd / 0x646D7067), 29 kb/s (default)
        Metadata:
        creation_time   : 2019-02-10 10:59:19
        handler_name    : GoPro MET  
        [...]        

Extracting the binary GPS data from MP4

With this data, we can create a binary file with the gpmd data inside. The following command copy the stream 0:3 from the file GH010039.MP4 as raw, and stores it in GH010039.bin

%ffmpeg -y -i GH010039.MP4 -codec copy -map 0:3 -f rawvideo GH010039.bin

The binary looks like:

00000000: 44 45 56 43 00 01 14 A4 44 56 49 44 4C 04 00 01    DEVC...$DVIDL...
00000010: 00 00 00 01 44 56 4E 4D 63 01 00 0B 48 65 72 6F    ....DVNMc...Hero
00000020: 37 20 42 6C 61 63 6B 00 53 54 52 4D 00 01 05 AC    7.Black.STRM...,
00000030: 54 53 4D 50 4C 04 00 01 00 00 00 DB 53 54 4E 4D    TSMPL......[STNM
00000040: 63 01 00 0E 41 63 63 65 6C 65 72 6F 6D 65 74 65    c...Acceleromete
00000050: 72 00 00 00 53 49 55 4E 63 04 00 01 6D 2F 73 B2    r...SIUNc...m/s2
00000060: 53 43 41 4C 73 02 00 01 01 A2 00 00 4D 54 52 58    SCALs...."..MTRX
00000070: 66 24 00 01 00 00 00 00 00 00 00 00 3F 80 00 00    f$..........?...
00000080: 00 00 00 00 BF 80 00 00 00 00 00 00 3F 80 00 00    ....?.......?...
00000090: 00 00 00 00 00 00 00 00 4F 52 49 4E 63 01 00 03    ........ORINc...
000000a0: 59 78 5A 00 4F 52 49 4F 63 01 00 03 5A 58 59 00    YxZ.ORIOc...ZXY.
000000b0: 41 43 43 4C 73 06 00 DB FD 35 FF 2D 0D C7 FD 39    ACCLs..[}5.-.G}9
000000c0: FF 19 0D 87 FC CD FE F9 0D 4B FC 05 FE 45 0C AF    ....|M~y.K|.~E./
000000d0: FB B5 FD ED 0C 8F FB 65 FD E1 0C 6B FB 11 FE 6D    {5}m..{e}a.k{.~m
000000e0: 0C 47 FA E5 FE ED 0C 53 FA C5 FF 89 0C 87 FA C1    .Gze~m.SzE....zA
000000f0: FF F9 0C B3 FA AD 00 9F 0D 33 FA A9 00 B3 0D 6F    .y.3z-...3z).3.o
00000100: FA 95 00 B7 0D BB FA 65 00 C7 0E 3B FA 45 00 D7    z..7.;ze.G.;zE.W
00000110: 0E 73 FA 49 00 EF 0E A3 FA 25 00 E7 0E FB F9 FD    .szI.o.#z%.g.{y}
00000120: 00 E3 0E FF F9 E9 00 F3 0E FF F9 E9 00 E3 0E F3    .c..yi.s..yi.c.s
[...]

Testing

Sample videos are downloaded from here. I will try to put an original Gopro7 file later. The gps data is extracted from the .MP4 file. The gpx, kml and bin files are stored in the repo. Karma introduces a new way of read GPS information and time, based on SYST and GPRI labels. I did my best trying to parse it. Seems accurate. If you have some long files to do, please extract the raw data and send me them (see extracting data).

  • fusion Fusion
  • hero5 Fusion
  • hero6 (all the points) Fusion
  • hero6 (only GPSFIX!=0) Fusion
  • karma Fusion
  • Gopro7 Gopro7

Status and future work

Currently, gopro2gpx generates hard-formatted kml, and a useful gpx file. But:

  • Not all tags are parsed. see fourCC.skip_labels and fourCC.labels for more info.
  • The are a little error detecting code.
  • Karma drone GPS info has been infered from debug. Maybe the SYST label is parsed wrong.
  • UNIT labels are parsed hardcoded.
  • Need ffmpeg and ffprobe to extract the data.
Owner
Chris Auron
Chris Auron
Tool to display your current position and angle above your radar

🛠 Tool to display your current position and angle above your radar. As a response to the CS:GO Update on 1.2.2022, which makes cl_showpos a cheat-pro

Miko 6 Jan 04, 2023
Streamlit Component for rendering Folium maps

streamlit-folium This Streamlit Component is a work-in-progress to determine what functionality is desirable for a Folium and Streamlit integration. C

Randy Zwitch 224 Dec 30, 2022
A proof-of-concept jupyter extension which converts english queries into relevant python code

Text2Code for Jupyter notebook A proof-of-concept jupyter extension which converts english queries into relevant python code. Blog post with more deta

DeepKlarity 2.1k Dec 29, 2022
Geodata extensions for Django REST Framework

Django-Spillway Django and Django REST Framework integration of raster and feature based geodata. Spillway builds on the immensely marvelous Django RE

Brian Galey 62 Jan 04, 2023
A Python framework for building geospatial web-applications

Hey there, this is Greppo... A Python framework for building geospatial web-applications. Greppo is an open-source Python framework that makes it easy

Greppo 304 Dec 27, 2022
A part of HyRiver software stack for handling geospatial data manipulations

Package Description Status PyNHD Navigate and subset NHDPlus (MR and HR) using web services Py3DEP Access topographic data through National Map's 3DEP

Taher Chegini 5 Dec 14, 2022
Implementation of Trajectory classes and functions built on top of GeoPandas

MovingPandas MovingPandas implements a Trajectory class and corresponding methods based on GeoPandas. Visit movingpandas.org for details! You can run

Anita Graser 897 Jan 01, 2023
Geospatial Image Processing for Python

GIPPY Gippy is a Python library for image processing of geospatial raster data. The core of the library is implemented as a C++ library, libgip, with

GIPIT 83 Aug 19, 2022
Deal with Bing Maps Tiles and Pixels / WGS 84 coordinates conversions, and generate grid Shapefiles

PyBingTiles This is a small toolkit in order to deal with Bing Tiles, used i.e. by Facebook for their Data for Good datasets. Install Clone this repos

Shoichi 1 Dec 08, 2021
A public data repository for datasets created from TransLink GTFS data.

TransLink Spatial Data What: TransLink is the statutory public transit authority for the Metro Vancouver region. This GitHub repository is a collectio

Henry Tang 3 Jan 14, 2022
Manage your XYZ Hub or HERE Data Hub spaces from Python.

XYZ Spaces for Python Manage your XYZ Hub or HERE Data Hub spaces and Interactive Map Layer from Python. FEATURED IN: Online Python Machine Learning C

HERE Technologies 30 Oct 18, 2022
🌐 Local tile server for viewing geospatial raster files with ipyleaflet or folium

🌐 Local Tile Server for Geospatial Rasters Need to visualize a rather large (gigabytes) raster you have locally? This is for you. A Flask application

Bane Sullivan 192 Jan 04, 2023
A bot that tweets info and location map for new bicycle parking added to OpenStreetMap within a GeoJSON boundary.

Bike parking tweepy bot app A twitter bot app that searches for bicycle parking added to OpenStreetMap. Relies on AWS Lambda/S3, Python3, Tweepy, Flas

Angelo Trivisonno 1 Dec 19, 2021
Geocode rows in a SQLite database table

Geocode rows in a SQLite database table

Chris Amico 225 Dec 08, 2022
Calculate the area inside of any GeoJSON geometry. This is a port of Mapbox's geojson-area for Python

geojson-area Calculate the area inside of any GeoJSON geometry. This is a port of Mapbox's geojson-area for Python. Installation $ pip install area U

Alireza 87 Dec 14, 2022
PyTorch implementation of ''Background Activation Suppression for Weakly Supervised Object Localization''.

Background Activation Suppression for Weakly Supervised Object Localization PyTorch implementation of ''Background Activation Suppression for Weakly S

34 Dec 27, 2022
Manipulation and analysis of geometric objects

Shapely Manipulation and analysis of geometric objects in the Cartesian plane. Shapely is a BSD-licensed Python package for manipulation and analysis

3.1k Jan 03, 2023
Minimum Bounding Box of Geospatial data

BBOX Problem definition: The spatial data users often are required to obtain the coordinates of the minimum bounding box of vector and raster data in

Ali Khosravi Kazazi 1 Sep 08, 2022
ProjPicker (projection picker) is a Python module that allows the user to select all coordinate reference systems (CRSs)

ProjPicker ProjPicker (projection picker) is a Python module that allows the user to select all coordinate reference systems (CRSs) whose extent compl

Huidae Cho 4 Feb 06, 2022
A NASA MEaSUREs project to provide automated, low latency, global glacier flow and elevation change datasets

Notebooks A NASA MEaSUREs project to provide automated, low latency, global glacier flow and elevation change datasets This repository provides tools

NASA Jet Propulsion Laboratory 27 Oct 25, 2022