Mapillary Downloads and GPS Coordinates?

Hello,

I have used Mapillary quite a bit in the past using the previous API. However, I am new to using the v4 API.

Happily, I have been able to download images using the Python code kindly provided here:

However, the images do not seem to contain the GPS coordinates. Is there a way to download the images with the GPS coordinates or can they be added to the images at a later stage?

Grateful for any suggestions, please.

Cheers.

However, the images do not seem to contain the GPS coordinates. Is there a way to download the images with the GPS coordinates or can they be added to the images at a later stage?

The short answer is that image metadata and image data are decoupled by design. And yes, you can put them together if you want to but it is your job to do so.

There are three basic technical and security reasons for this:

  1. Implementing a mapping portal that is easy to navigate and develop for is simpler this way
  2. Managing huge amounts of data and metadata in separate data silos is more efficient and more effective
  3. Serving imagery deprived of metadata is one safeguard against proliferation of data which could potentially leak private data, or data which could be used to trace individuals (when put together with other sources), especially when shared in the digital age, namely on the internet.

I have used Mapillary quite a bit in the past using the previous API.

Then you surely know that the principles stated above have applied then and still apply today. Is it perfect? No, but we should not make life easier for malicious actors than necessary nor should we endanger well minded users who “just want to share an image on the internet” to unwittingly proliferate possibly personal data.

Hi, thanks for the reply!

And yes, you can put them together if you want to but it is your job to do so.

Is there an efficient way to match the GPS coordinates to the images once they have been downloaded?

Much appreciated.

Is there an efficient way to match the GPS coordinates to the images once they have been downloaded?

What do you mean by efficient? You simply download images and store them by their key names, then you iterate over the keys and add GPS*‑Exif tags to the image files. Done. Iteration is the most efficient control flow type of operation on John von Neumann machines.

Yes, that’s exactly what I meant. Do you know if there is any code available that would do this, please?

I found a relevant post but it uses GeoSetter:

Cheers.