Real-Time classification of Plankton species using Convolutional Neural Networks

Marine geology involves investigation of marine species for ocean observations. Marine geologists study the history, the processes occurring at the ocean floor to derive valuable insights about marine life. This study plays a vital role in maintaining balance in our ecosystem. The marine ecosystem houses species ranging from energy producers such as aquatic plants, phytoplanktons and consumers like fishes to humans. The study of one such marine species called "plankton" is known as planktology. Planktons are microscopic species that play a rather unnoticed, but a pivotal role in the marine food web. This leads to a need for a proper identification system which classifies different plankton species and also has a record of them which further helps in recognizing new plankton species. This system can also be used as a tool for classification of a particular plankton species with the help of Convolutional Neural Networks, which is a tedious process otherwise. This model is further deployed as a web app which can also allow authenticated users to contribute to the data collection of the model. Thus, this system will have a major impact on knowing different plankton species and also makes the subject of planktology more feasible.