This project is part of the joint internship program with PT Jala Akuakultur Lestari Alamku, Department of Electrical Engineering and Information Technology Universitas Gadjah Mada (DTETI UGM), and Toyohashi University of Technology (TUT). In a 3-month timeline, each small team is assigned to a different role with different projects. The expected output is an usable prototype and an internship report.

I am placed on a software engineering team consisting of 4 people. Our objective is to create a module that parses a handwritten log by shrimp farmer into records in the database. We start by survey and collecting real data from shrimp farmers. Then use the collected data to experiment the approach. We also annotate the data to enable quantitative metrics. Finally, we successfully built a prototype web app according to requirements and presented our work with the JALA team.

In this team, I am responsible for the algorithm and approach to “parse” the data records from OCR reading. Unfortunately, I could not specify further about the details of implementation because of NDA. However I am allowed to say that I created a module using native JavaScript for the initial prototyping that uses a lot of geometry to estimate the table contour and data location.