Problem Statement 4: Our challenge for the fishackathon is to develop a smartphone app that captures data on fishing practices and catch composition with the least amount of manual data entry, and with feedback to the fisher, in the same spirit as activity trackers such as Strava, Runtastic, or Endomondo.
What it does
Our app allows users of all kind, from commercial to recreational, grandparents to kids, to capture images and data of their catches. The app automatically generates the length of the fish as well as the location it was caught. It is then saved to our database, where it is readily available to researchers for further analysis.
This app is useful for both users and the scientific community. For the users, this app provides a quick, easy, and fun way to keep a log of their catches and share their fishing accomplishments on social media. Being able to see the length also adds a sense of competitiveness, further encouraging users to use the app. For researchers, extracting valuable length and geolocation data from something users are already doing, taking pictures, will significantly increase the amount of data available.
How we built it
The core feature of this app is its innovative length calculation algorithm. By detecting subtle changes in your phone camera's focal distance as takes the picture, we are able to calculate the length of the object to a high degree of accuracy. All in the press of a button. And the best part? All you need is a smartphone -- no size reference required!
Accomplishments that we're proud of
Innovative way of measuring length with phone camera, without the need for special hardware or a backdrop.
Challenges we ran into
We forgot to bring a ruler. There were no rulers at fishackathon.
What we learned
It turns out, when you're trying to build a length-measuring app, a ruler is kinda handy.
What's next for Fishing for Data
We are saving as much data as we can (pictures of fish, user-inputed species) in hopes that software like FishFace will one day be able to detect the species of fish from our pictures. This will provide users more information about their catch (and hopefully make them more aware of sustainable fishing practices) and allow researchers to determine valuable data about each species, such as length, location, and catch rate. We are also hoping that developers of software such as FishFace will find our user-inputed species information useful in training their fish species classifiers.