Project overview

A smart solution with an ability to swiftly recognize parking lots from satellite images for effective car park management.

Delivery Model
SLA-driven task backlog delivery (TM-based)
Effort and Duration
1 month, 2 man-months
tensorflow.js, electron.js, retinanet

85% recognition accuracy

  • Parking lot detection based on Google Maps
  • Easy compilation from a single codebase for any platform to save time and effort
  • In-browser object detection using local resources
  • Speed-optimized supervised deep learning algorithms
  • Exportable image recognition results
  • Fast system training for new use cases