This project can be split into 3 main parts.

Part 1) Location of a suitable area of ocean to search for marine debris over a suitable time frame (3-4 years). This will have to consider year-round cloud cover and sea state, as well as the degree to which debris is expected to aggregate in the area.

Part 2) Collecting and filtering a satellite dataset to account for cloud cover and sea state. Train a machine learning model that can identify marine debris from satellite images. This could be adjusted to focus specifically on plastics.

part 3) Develop an application that incrementally gathers suitable satellite data, uses the machine learning model to classify debris, and creates a density map of plastic hotspots