17:15 〜 18:45
[AOS11-P02] Research of Species Identification for the Establishment of “Core Ocean” in Space
キーワード：コアバイオーム、海洋生態系、サンゴ
Currently, projects such as the Artemis Project aimed towards migration to Mars and the Moon are being developed. Among these, Yamashiki et al. have proposed the "Three Core Concepts for Space Migration," aimed at establishing a core academic system for space migration, with the assumption that human migration to the Moon and Mars becomes a reality in the latter half of the 21st century. The "Three Core Concepts for Space Migration" consist of a "core biome complex", “core technologies", and “core societies". “Core biome complex” is defined as the ecological systems on Earth extracted and classified by function and nature, to create a "selected core biome" containing elements deemed necessary for migration. In particular, the Core Biome aims to address the question of whether humans can survive in an artificially constructed ecosystem and defines "Core Ocean" and "Core Forest" as part of this goal. In the Core Ocean, the objectives are further divided into two categories, the selection and breeding of species that will become the core of marine ecosystems in the closed environment of space aquaculture. The Core Ocean absorbs carbon dioxide from the atmosphere, which in space can be fixed by productive activities and create a stable environment. Furthermore, the Core Ocean is home to a great variety of edible fish species serving as a potential food sources.
Corals, being cnidarians and capable of photosynthesis, help stabilize the concentration of carbon dioxide in the ocean. In addition, coral reefs occupy only 0.2% of the total area of the ocean but are home to about 1/3 of marine life, allowing them to function as a central element in a complex ecosystem. The goal of this study is to attempt to identify species of coral which can be used in the selected core biome. This is done by breeding and maintaining the coral and its surrounding marine organisms to observe and study the formation of a core ocean centered on coral in the closed environment of space.
To form a Core Ocean based on corals, it is necessary to select species that can withstand changes in water composition which can occur in a closed ecosystem in space. In addition, corals serve as habitats for marine organisms, making the selection and monitoring of fish species living in the surrounding coral an extremely important point. The monitoring of the fish and corals must be done through video to intuitively determine the distance between individuals and their behavior. This can be done through a system that allows species identification in videos taken during dives and confirming the results with real-time video discrimination. For this, we used YOLO, an object detection application that specializes in real-time video discrimination. In this study, we created our data set by taking video of coral and fish inhabiting the Ishigaki and Iriomote Islands, Okinawa Prefecture, where coral reefs are stably developed and can be photographed year-round, and in Tanabe City and Kushimoto Town, Wakayama Prefecture in summer. Currently, object detection is focused on fish, with the goal of building a system that can identify and monitor coral as well.
Corals, being cnidarians and capable of photosynthesis, help stabilize the concentration of carbon dioxide in the ocean. In addition, coral reefs occupy only 0.2% of the total area of the ocean but are home to about 1/3 of marine life, allowing them to function as a central element in a complex ecosystem. The goal of this study is to attempt to identify species of coral which can be used in the selected core biome. This is done by breeding and maintaining the coral and its surrounding marine organisms to observe and study the formation of a core ocean centered on coral in the closed environment of space.
To form a Core Ocean based on corals, it is necessary to select species that can withstand changes in water composition which can occur in a closed ecosystem in space. In addition, corals serve as habitats for marine organisms, making the selection and monitoring of fish species living in the surrounding coral an extremely important point. The monitoring of the fish and corals must be done through video to intuitively determine the distance between individuals and their behavior. This can be done through a system that allows species identification in videos taken during dives and confirming the results with real-time video discrimination. For this, we used YOLO, an object detection application that specializes in real-time video discrimination. In this study, we created our data set by taking video of coral and fish inhabiting the Ishigaki and Iriomote Islands, Okinawa Prefecture, where coral reefs are stably developed and can be photographed year-round, and in Tanabe City and Kushimoto Town, Wakayama Prefecture in summer. Currently, object detection is focused on fish, with the goal of building a system that can identify and monitor coral as well.