Skip to content

kamangir/vancouver-watching

Folders and files

NameName
Last commit message
Last commit date

Latest commit

ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 

Repository files navigation

๐ŸŒˆ Vancouver Watching (vanwatch)

vanwatch ๐ŸŒˆ discovers and ingests images from traffic cameras in an area and then runs YOLO ๐Ÿš€, OpenAI Vision, and other vision algo to extract information about urban activity at scale.

pip install vancouver-watching
 > vanwatch help
vanwatch conda create [validate,~recreate]
 . create conda environment.
vanwatch conda validate
 . validate conda environment.
vanwatch discover \
	[area=<area>,~upload] \
	[-|<object-name>] \
	[<args>]
 . discover area -> <object-name>.
vanwatch ingest \
	area=<area>,count=<count>,dryrun,gif,model=<model-id>,~process,publish,~upload \
	-|<object-name> \
	[<args>]
 . ingest <area> -> <object-name>.
vanwatch list [area=<area>,discovery|ingest,published] \
	[--count <count>] \
	[--delim space] \
	[--log 0] \
	[--offset <offset>]
 . list objects from area.
2 area(s): iran,vancouver
vanwatch list areas
 . list areas.
vanwatch vision "prompt" \
	[area=<area>,offset=<1>,auto|low|high,dryrun,~upload] \
	Davie,Bute \
	[--verbose 1]
 . openai_commands vision: prompt @ <area>/intersection.
vanwatch process \
	count=<count>,~download,gif,model=<model-id>,publish,~upload \
	.|<object-name> \
	[--detect_objects 0] \
	[--do_dryrun 1] \
	[--overwrite 1] \
	[--verbose 1]
 . process <object-name>.
vanwatch pylint
 . pylint vancouver_watching.
vanwatch update|update_cache \
	area=<vancouver>,overwrite,process,~publish,refresh,~upload \
	[--verbose 1]
 . update QGIS cache.
vancouver_watching test \
	[dryrun,~ingest,~list,~process,upload]
 . test vancouver_watching.

last build ๐Ÿ”— image

discover and Ingest an Area

image

to see the list of areas supported by vanwatch type in,

vanwatch list areas

to discover the available cameras in an area type in,

vanwatch discover area=vancouver

you have generated a geojson of traffic images in the City of Vancouver. Now, you can ingest the traffic images from this area and detect people and cars in them,

vanwatch ingest area=vancouver,count=2,publish

image

model: https://hub.ultralytics.com/models/R6nMlK6kQjSsQ76MPqQM?tab=preview

image

image

image

dataset: vanwatch-cache-2024-02-28-21-04-19-26236.tar.gz (details).


PyPI version

to use on AWS SageMaker replace <plugin-name> with vanwatch and follow these instructions.

About

๐ŸŒˆ Vancouver watching with AI.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published