Description
The Idea
Currently, users have to take a photo and upload it. It would be much faster if they could open a live camera feed inside the browser, point it at a fish, and have the app automatically capture a frame when the fish is perfectly in focus.
What needs to be done
- Use the HTML5
MediaDevices API to stream the camera feed directly into a <video> element on the Scanner page.
- Add a "Capture" button that grabs the current frame and sends it to our ML backend.
- (Bonus) Draw a framing guide (a subtle bounding box) over the camera feed.
Description
The Idea
Currently, users have to take a photo and upload it. It would be much faster if they could open a live camera feed inside the browser, point it at a fish, and have the app automatically capture a frame when the fish is perfectly in focus.
What needs to be done
MediaDevicesAPI to stream the camera feed directly into a<video>element on the Scanner page.