Skip to content

Latest commit

 

History

History
11 lines (6 loc) · 898 Bytes

RealtimeObjectDetection.md

File metadata and controls

11 lines (6 loc) · 898 Bytes
Raw

Realtime Object Detection

alt text

This is an Intelligent Edge scenario built on top of Windows ML and ONNX to bring to life a completely offline object detection experience that runs directly from the camera feed. This demo uses a model built and exported by Microsoft Custom Vision.

Key Source Code

  • RealtimeObjectDetection: Main page that drives the demo. It hosts the CameraControl (see below) to display the live camera feed, creates a background loop to capture and process frames from the camera. It then uses the result to detect objects in the image.

  • CameraControl: The code that contains the camera feed and enables camera captures.