Jaime Jacob
Pronouns: she/her
Research Mentor(s): Katelyn King
Research Mentor School/College/Department: / SNRE
Program: UROP
Session: Session 4 (1:40pm – 2:30pm)
Authors: Jaime Jacob, Katelyn King, Karen Alofs
Abstract
Fish communities in Michigan inland lakes over the last century have been managed for various reasons, such as controlling fish populations, providing ecosystem balance, and increasing sportfishing opportunities (Whelan, 2004). These management strategies include habitat management, species management with stocking and eradication, and fishing regulations. However, in the 20th century, management recommendations and actions for each lake were recorded on paper cards and stored for future reference. In order to be used for current fisheries research and management, the data on these cards needs to be in a digital and accessible format. Therefore, our goal was to digitize these records so that they could be to investigate the impacts of historical management actions. In this project, we transcribed JPEG images of these cards to comma-separated values (CSV) files using optical character recognition (OCR) technology. Amazon Web Services were used to organize the 5500+ images and then run them through the OCR program. This process resulted in a CSV file for each image that we initially cleaned by writing code using Python programming. Finally, the files were manually compared to the original JPEG image to ensure accuracy and consistency. As a result of this process, 5500+ images will be transcribed and stored in an effective manner for easy access and use. With this database configured, the information from the historical records can be utilized for research which will inform future management.
