Plastic pollution has become a significant issue facing today’s oceans. This contamination is complicated further by the presence of small particles (including microplastics), which can either be released directly into the environment or broken down from larger plastic components. A promising, lower-cost approach for detection of microplastics involves the implementation of the Nile Red dye. This ongoing research presents a design seeking to address some of the existing complications with Nile Red as a low-cost method for in situ microplastic detection. This approach implements a machine learning based fluorescence imaging approach. Nile Red dye is able to stain plastics and fluoresce under blue light, making plastics more easily identifiable to a standard camera. Plastics were left to stain in a dark room for 30 minutes in a mixture of 3 μg/mL of Nile Red dye. These plastics were then moved to a tubing setup connected to a reversible pump, which oscillated between moving the plastics in each direction. This tubing setup was then lit by blue LEDs, and recorded by a phone camera using an orange filter to block out the blue light while allowing through the fluorescence spectra. These tests were done with different known masses of plastics, with each test being run for one minute at 240 frames per second. The frames of the video were run through an image labeler, using a script to detect plastic pieces based on their colors and shape. This was then used as the data set to detect any plastics in frame. In order to detect the plastic concentration in a video, the presence of microplastics was detected in each frame using a deep learning algorithm trained on the plastic image data set. The total pixel area covered by microplastics across the entire video was calculated, and then averaged over the total pixel area of the entire video. This could then be plotted against videos of known microplastic concentrations, allowing for new videos to be analyzed and receive a microplastic concentration estimate. Although still ongoing, this project has the potential to expand the possibilities for lower cost microplastic analysis.