A capture device is passing through space and an AI system is segmenting the data in the file (e.g., the doors in a hall, the houses on a road, furniture in a room) and the application layer seeks to extract information in a consistent, repeatable way the features of interest (e.g., the doors, the houses, the tables, etc).
Howard comment: didn't we already do this in CityGML? Has anyone looked at the spec from that perspective?
The entire structure of a building is defined as a collection of other things (e.g. doors, windows, objects in spaces).
Show me Howard's door:
could be parsed as "show me all of Howard's doors"
alternatively could be parsed find Howard's cubicle and then find the door for that cubicle
In first case, it's a contained relationship: Howard's doors is a feature class. Contains all of Howard's doors
In second case, I must first have Howard's cubicle, then the door is a feature of that.
This should have been a use case in the list of use cases we brainstormed. That's where this belongs.
A capture device is passing through space and an AI system is segmenting the data in the file (e.g., the doors in a hall, the houses on a road, furniture in a room) and the application layer seeks to extract information in a consistent, repeatable way the features of interest (e.g., the doors, the houses, the tables, etc).
Howard comment: didn't we already do this in CityGML? Has anyone looked at the spec from that perspective?
The entire structure of a building is defined as a collection of other things (e.g. doors, windows, objects in spaces).
Show me Howard's door:
In first case, it's a contained relationship: Howard's doors is a feature class. Contains all of Howard's doors In second case, I must first have Howard's cubicle, then the door is a feature of that.
This should have been a use case in the list of use cases we brainstormed. That's where this belongs.