Questions & Answers - Detecting informal settlements via topological analysis

Disclaimer: This is an archived version of the once interactive session pad with content from conference attendees. Please note that some information was lost during the transformation (i.e. the edit history, user colours, the chat and some formatting).

Back to session page

Detecting informal settlements via topological analysis Abstract in the proceedings of the SotM 2020 Academic Track: https://zenodo.org/record/3928690 Full proceedings of the SotM 2020 Academic Track: https://zenodo.org/communities/sotm-2020

Questions

  1. [ANSWERED] Can you comment on informal settlements detected in South Africa? asseblief [please] ;-) – Serena. Thanks, we are mapping away, but it may take some time…
    1. (comment: https://www.openstreetmap.org/#map=17/-34.06021/18.66231,https://millionneighborhoods.org/#17/-34.06021/18.66231  – YaguraStation)
  2. [ANSWERED] Maybe off-topic, regarding above example in 1. (Monwabisi Park): City planning-wise how would their situation be improved(eg. in regard to accessibility)?
  3. [ANSWERED] You discussed reblocking, and in your paper you show how “We observe that new infrastructure segments typically appear as dead-end streets (culs-de-sac), as the minimal edge set needed to connect a collection of nodes will always yield a tree graph.” But a street network of dead-end streets is notoriously harmful to walkability (as in USA suburbs), and will result in residents having to take long network-distance walks when the absolute distance of their trip is short. How can you help ensure this will not be used to design neighborhoods that are difficult to walk in? Thanks for your fascinating talk! -Taylor, ITDP (thanks for reading the long question!)
    1. Marco: you are welcome, thanks for the interest!
  4. [ANSWERED] Is it possible to assess/compare the quality of the settlements with the OSM boundaries?
  5. [ANSWERED] What are the software tools you used to perform the work/run the model? Is the source code available? – Marco Minghini

Comments

  1. [ANSWERED] If those “cadastral maps” ever leaked to the commumity they would go nuts: “Hey that’s my land. You gave it to neighbor C”, etc.-jidanni
    1. Even real cadastral maps are sensitive. (usually due to not being aligned precicsly… even that is source of endless arguments.)-jidanni
    2. Even slightly out of date real cadastral maps are destroyed by govenments, to avoid misunderstandings.
    3. No matter if there is big warning “not a real cadastral map”… people will take it as the real thing.-jidanni
  2. Yes I wanted to do a historical cadastral map project… only to find even the parcel numbers had been erased by the Taiwan government due to “sensitivity”. (Could only query one by one.)
  3. All the buildings I trace from imagery are squwed several random meters thanks to Bing, etc. I hope you have better grasp of real WGS84 locations.-jidanni

OSM Science Mailing list: https://lists.openstreetmap.org/listinfo/science