Processing

Intel, Microsoft Azure, Amazon Web Services (AWS), and several other large tech giants provide software tools for analyzing imagery using Convolutional Neural Networks (CNN). This is a form of machine learning and Artificial Intelligence (AI) that can streamline the processing of refining drone data. Object-Based Image Analysis (OBIA) is commonly used for processing the very high-resolution images collected by UASs (Blaschke et al., 2014). If the program can be trained on what seagrass, deer, or geese look like in the imagery, the program can automatically detect the presence of these variables, and provide a total count through the image recognition process and statistical analysis. Then, the count can be updated periodically as conditions change, and the total increase or decrease in abundance can be shown in an easy to understand graph or chart with percent certainty.

This is an important step in gathering data that will be shared with those who use it to craft rules and regulations that protect wildlife in and around the Chesapeake Bay. This same method can be used for tracking algal blooms. This will help identify areas that need more attention so that scientists can identify the source of the problem before its effects become irreversible.