Travel-Lapse

Data Collection

Using GPSLogger for Android (available on F-Droid) to collect CSV files with geo-locations. Create daily zip files, which are uploaded into the /gps folder. (The /locations folder contains data from the now mostly defunct Google Location History.)

Scripts

The scripts form a kind of pipeline, each creating files required for the next script.

There is a script called run_all_scripts.py which runs all the other scripts. Ideally, this just works.

All scripts have an individual overwrite variable, which by default is set to False. It controls whether files are recreated or if only files are created that do not yet exist.

Folder Structure

Not sure all scripts create the folders they need... hopefully. Anyway, the structure looks like this:

├── all
├── basisdaten
├── csv
├── cumulative
├── cumulative visualisation
├── fast
├── gps
├── locations
├── output
├── points
├── points_yearly
├── shapefile_cumulative
├── shapefile_yearly
├── slow
├── static
    ├── droid
    ├── fantasque
    ├── jetbrains
    ├── mono
├── venv
├── visualizations
├── visualizations_geopandas
├── visualizations_square
├── visualizations_square_with_dates
├── visualizations_vertical
├── visualizations_vertical_with_dates
├── visualizations_with_dates
├── visualizations_with_years
├── visualizations_yearly_geopandas
All the Python files are in the root.

prepare data

To start the process there needs to be CSV Files, with the location track for a day, in the /csv Folder, with a naming pattern like yyyymmdd.csv. These scripts might help you create them.

• extract_csv_files.py extracts the data from /gps and /locations into CSV files, creating one CSV file with geo-locations per day. Preferably, it uses a copy from the CSV in a zip file in /gps. Only if this does not exist does it go to other sources. At the end, there should be a CSV file with a bunch of geo-locations for each day, using a yyyymmdd.csv naming scheme like this:

time,lat,lon,elevation,accuracy,bearing,speed,satellites,provider,hdop,vdop,pdop,geoidheight,ageofdgpsdata,dgpsid,activity,battery,annotation,timestamp_ms,time_offset,distance,starttimestamp_ms,profile_name,battery_charging
2024-03-19T23:12:40.600Z,48.1861084,11.5593367,561.6000366210938,,,,,network,,,,,,,,,,,,,,,

• split_csv_by_day.py is an alternative to the script above. It takes CSV files in the format described above from the folder /csv_raw and splits same into one file per day into /csv. This might be helpful if your CSV data is organized monthly.

create images and videos

• calculate_speed_and_filter.py takes the created CSV files and calculates the speed between two points. It then creates four GeoJSON files: one is a line between all the points of a day in the folder /all, one only contains lines if the speed between those points is above 10 km/h (/fast), the next one only contains lines between points below 10 km/h (/slow), and last but not least, /points contains points every 500 meters along the lines with a speed below 10 km/h.

• cumulative_points.py takes the points from /points and creates a cumulative points file in /cumulative with a naming scheme like this: 20200319_points.geojson. These include all points up to that date. Even if no location file exists for a day, a cumulative one is still present. From now on, every date from the start date is covered. You need to set the start date in the header of this file!

• combine_points_yearly.py takes the points from /points (which only include points for distances traveled at less than 10 km/h) and creates a file for each year.

• visualize_points_with_counts.py creates shapefiles for the yearly points created with combine_points_yearly.py.

• visualize_cumulative_points_with_counts.py takes the files created in cumulative_points.py and creates a shapefile with the counts of the points in each polygon. It creates one shapefile for each day.

• visualize_points_geopandas.py takes the shapefiles created in visualize_cumulative_points_with_counts.py and creates a .png image using background data from the /basisdaten folder for each day. You need to set the start date in the header of this file!

• visualize_points_geopandas_yearly.py creates a yearly visualization and an additional file that writes the name on it.

• create_cropped_images.py creates cropped images (square and vertical) of the images created by visualize_points_geopandas.py.

• create_video_from_images.py creates a copy of each of the .png files created by visualize_points_geopandas.py and adds the date to the lower right corner (/visualizations_with_dates). It then also crops these into square and vertical images and adds the date to those as well. All these images are then combined into three .mp4 files (16:9 4K, vertical, and square video).

Example: https://www.youtube.com/watch?v=zHYTjOnBznY

Note on Use of LLMs

This code was created using, among other tools, LLM tools like ChatGPT.