Skip to content
Zeynoc
de
All work

Case study · Mussel App by FutureLab

Measuring mussels from a phone photo

An image-processing API for Mussel App by FutureLab. A calibrated photo becomes measurements the team can review in millimetres.

musselfarm.co.nz
Millimetres from one photo
for every mussel on the plate
No special camera
a phone photo with a printed scale
Runs as an API
used by the team’s iOS app

The challenge

Mussel App is farm management software for New Zealand mussel and oyster farmers. Harvested mussels were measured one by one with callipers. It was slow, and two graders would often get different results.

The goal

Turn one phone photo of mussels on a measuring plate into length and width in millimetres, reliable enough for grading and simple enough for use on a boat.

What we built

  1. A YOLO model finds and outlines every mussel in the photo.
  2. A second model reads the ruler printed on the plate, so pixels convert to millimetres on every image, at any distance.
  3. Each shell is measured and the results are drawn back onto the photo for a quick check.
  4. Everything is served through a FastAPI service on AWS Lambda, ready for the iOS app.

The hard parts

Photos taken from different distances.

Calibration from the plate's own scale in every photo.

Phone rotation and EXIF quirks.

Automatic orientation correction before any measurement.

Busy steel benches in the background.

Directional filtering to isolate the ruler lines.

Blurry or partly hidden shells.

A fallback that still estimates size and marks low confidence results.

The result

Consistent, hands free mussel sizing in seconds, replacing manual calliper work for the team.

What would you like to build?

Tell us where your project stands. If AI is not worth it for you, we will say so.

hello@zeynoc.com