Improve Document Scanning: Edge Detection and Optimization

Yeah I think the algorithm that draws the bounding box when scanning documents using the phone camera could be improved. It's slow to recognize the edges, shows a bounding box like 30 percent of the time, and then takes like 4 seconds to process the photo. I usually have to manually realign the corners on every scan I make. This is what I primarily use the mobile app for anyways. It's my number 1 used feature.

I like these current features:
1. Automatic edge detection and automatic capture (works once in a blue moon though)
2. Option to manually adjust corners
3. Button placement is consistent and in the right place for my thumb. Good choice

Room for Improvement:
1. Faster edge detection
2. More robust lighting conditions (with different temperatures of light)
3. The ability to name the document after pressing the save button. instead of hunting for the scan after its been uploaded to my files.
4. Smaller file sizes after saving
5. Filters to increase the clarity of handwritten lines, pen ink, pencil, etc.

I would like this mobile scanning feature to be equal to the ease of use and quality of output from the app Microsoft Lens (now defunct). I feel like its 70% of the way there.

No files, tests, or entry points are named. Start by locating the mobile document-scanning flow and reproducing edge detection, capture, and processing delays under varied lighting. Before changing code, narrow the request to one measurable improvement; done should include focused tests or benchmarks and a clearly agreed result.

Written by the indexing model from the issue text.