Picstonic App
A Flutter background-removal app built at Bixosoft as internal R&D. Pick or capture a photo and a Django service running OpenCV segmentation returns a cut-out with a transparent background. It stayed an internal build and was never taken to production.
What Problem Did This Project Solve?
Background removal on a phone, without shipping the processing to the device.
- Image segmentation is too heavy to run comfortably on the mid-range Android hardware this was aimed at, so the work had to happen off-device.
- Photos straight from a phone camera are large; uploading them unprocessed over mobile data makes an app feel broken before it has done anything.
- The user is waiting on both a network round trip and a CPU-bound job, so the interface has to communicate progress rather than appear frozen.
- The output is a PNG with an alpha channel, which has to be previewed, saved and shared correctly on both Android and iOS.
- Classical computer vision cuts cleanly around high-contrast subjects and struggles on fine edges such as hair, so the interface had to set honest expectations about what it could do.
What Solution Was Delivered?
A thin Flutter client over a Django processing service, with the wait made visible.
- Camera capture and gallery selection, with client-side resizing before upload to keep payloads small on metered mobile data.
- A typed API layer against the Django service so upload, processing and retrieval are one predictable flow with one error path.
- Explicit progress and failure states, because the user is waiting on a server-side job rather than an instant local operation.
- Transparent-PNG preview rendered over a checkerboard so the user can actually see what was cut out.
- Save-to-gallery and share handled per platform across Android and iOS.
What Can This Project Do?
Camera & Gallery Input
Capture a new photo or pick an existing one, with resizing applied before upload.
Background Removal
The subject is cut out server-side and returned as a PNG with a transparent background.
Visible Progress States
Upload and processing progress surfaced explicitly, so a slow job never looks like a frozen app.
Transparency Preview
Results previewed over a checkerboard so transparent regions are actually visible.
Save & Share
Export the cut-out to the device gallery or share it directly, handled per platform.
Django API Client
A typed client over the Django endpoint, with one predictable flow for upload, processing and retrieval.
Built With Modern Engineering Standards
Have a project requiring modern web architecture, mobile apps, or autonomous AI pipelines?
Let's discuss how we can engineer a custom solution tailored to your operational bottlenecks.