Beaches are places where human living areas, and a variety of litter are washed ashore from the river and offshore. Plastic litter comprises 70-90% of beached litter, and they are thought to be fractured into microplastics by the effect of ultraviolet, winds, and waves. Automated analysis methods for the beach images taken by multi different types of cameras (e.g. aerial drones, webcams, and smartphones) have the possibilities to provide alternative monitoring methods to relatively laborious approaches such as a manual collection and counting of the litter and help us to accelerate to understand the real state of the pollution. As the first step to establishing the method to estimate the total amount of macro plastic litter on the beach, we developed a deep learning model which gives pixel-level image classification to the beach images into eight classes, including artificial litter and natural litter classes. The accuracy of the model was evaluated qualitatively and quantitatively. The detection accuracy of the model for artificial litter on the test data was around 80%, and the usefulness of the method was demonstrated by comparing the results by projection transform from the ground image inference, and drone image ground truth results. Moreover, now we are attempting to develop a technology to classify macro plastic objects in detailed classes. The training dataset used in this research is now public from SEANOE - Sea Open Scientific Data Publication. We also produced a beach plastic litter training dataset for machine learning use.

Primary Presenter: Mitsuko Hidaka, Japan Agency for Marine-Earth Science and Technology (JAMSTEC) (mitsukou@jamstec.go.jp)

Authors:

Mitsuko Hidaka, Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Kagoshima University  (mitsukou@jamstec.go.jp)

Koshiro Murakami, Japan Agency for Marine-Earth Science and Technology (JAMSTEC)  ()

Daisuke Sugiyama, Japan Agency for Marine-Earth Science and Technology (JAMSTEC)  ()

Shintaro Kawahara, Japan Agency for Marine-Earth Science and Technology (JAMSTEC)  ()

Shin’ichiro Kako, Kagoshima University, Japan Agency for Marine-Earth Science and Technology (JAMSTEC)  ()

Daisuke Matsuoka, Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Kagoshima University  ()