| 英文摘要 |
The rising awareness of coastal wastes in recent years has stimulated the aim to understand the current total number of marine debris in Taiwan. Thus, more research methods on investigating coastal wastes were introduced, and the baseline data has been accumulated to achieve the goal of integrating the information platform gradually.
This Project has established the survey method on coastal wastes, including the types of wastes, plastic bottles survey, and the principle of estimating the weight of coastal wastes. In addition, this Project has requested the assistance of 19 coastal cities ＆ counties’ Environmental Protection Bureaus (EPB) to conduct a survey every season. To ensure the smooth implementation of the coastal waste survey, this Project has provided training and guidance to 200 people, who were mostly the staff of local EPBs and cleaning squadrons in the 19 counties and cities.
The research topic of this Project is using the Deep Neural Network (DNN) to identify plastic bottles. This year, 60,000 photos have been added to the training data. The primary identification rate is 20-60% which fluctuates based on the coastal area where the monitoring took place. Sandy lands lead to higher misreading due to the gravels and small crushed stones which further results in a lower identification rate. If more training data could be added to improve the identification rate, it is estimated that the amount of manpower and time could be saved.
Taiwan Environmental Protection Administration (EPA) promulgated the Action Plan of Marine Debris Governance in Taiwan in 2018. Within, the plan of “Identifying Priority Coasts” has taken literature from South Korea, Japan, and European Countries into consideration. Additionally, specialists consultation meetings were hosted to take environmental organizations, experts, and scholars opinions into account. The first version of the identification chart for “Identify beaches to be cleaned” is completed. Six measures are used for the assessment, and coastal cleanup priority is given to those with higher points.
Regarding the system, emphases were placed on optimizing the system’s functions and boosting the usage frequency by the general public. Firstly, the general public can now apply through the system to host a coastal cleanup event, and the application review function is also established. Secondly, a new homepage is now provided for various users. Thirdly, the map overlaying function in the Geographic Information System (GIS) has also been provided. Lastly, the functions to produce statistics reports and to announce the results of coastal cleanup have been established. The above-mentioned functions, including maintaining the existing functions and establishing new ones, have successfully increased the user-friendliness of the system, the management effectiveness and monitoring the current coastal businesses.
|