Parking Lot Counter

Status: Live.

1. Your muse

   Upload a photo of a parking lot and it counts the cars and the empty spots, and outlines each car.

2. Show thinking

3. Checks passed

4. 

   The plan: Parking Lot Counter

   Upload a lot photo to count cars, spot empties, and see outlines.

   1. Upload a photo of a parking lot
   2. Review auto detected cars and empty spots
   3. See counts and photo with each car outlined

   See the full plan

   Who it's for

   Drivers, store staff, and lot managers who want a quick count of parked cars and open spaces.

   How it works

   1. Upload a photo of a parking lot
   2. Review auto detected cars and empty spots
   3. See counts and photo with each car outlined

   Screens

   • Upload. Upload a parking lot photo and start analysis.
   • Results. Show car and empty spot counts with outlined photo.

   What goes in

   • Image: One JPG or PNG parking lot photo. Daylight overhead or angled view with visible cars and lines works best.

   What comes out

   • Overlay: Same photo with each car outlined and numbered.
   • Text: Counts for parked cars, empty spots, and total spaces.
   • JSON: List of car boxes and spot counts for reuse.

   Models

   Model	Used for	Why
   SAM 3.1	Find and outline every car in the photo	It returns boxes and masks for cars from a short prompt like car.
   Muse Spark 1.3	Count empty marked spots and confirm totals	It reads the photo and lot markings to estimate free spaces and return structured counts.

   Limits

   • Counts are estimates. Cars hidden by trees, shadows, or other cars can be missed.
   • Empty spots are inferred from visible painted lines and clear pavement. Unmarked or cropped areas are not counted.
   • Works on one photo up to 4.5 MB via temp upload. No live video or lot maps.

   From your inspo

   The person gave a text note with no images. They want a tool that counts cars and empty spots from a parking lot photo and outlines each car.

   Takes

   • counting parked cars from a photo
   • counting empty parking spots
   • outlining each car

   Cost per run: about $0.007 (est.). SAM 3.1 about $0.002 per image plus Spark about $0.003 input and $0.002 output. Total about $0.007 per run.

6. Done. npm run verify exits 0 — all six steps pass, including both Playwright tests with real model calls.

   It's live.

   musemuse-parking-lot-counter.vercel.app (opens in a new tab)

   Product pageNew muse

   5 m 35 s · $0.54 est.