Identification and Quantification of Rockweed Biomass using High Resolution Satellite Imagery

阿卡迪亚海洋植物有限公司（ASL）是全球公认的食品、生化、农业和农化市场的海藻产品加工行业领导者，并为全球食品市场种植和加工独特的海藻。海藻Ascophyllum nodosum （Rockweed）是生产产品的主要原料。最终产品销往全球70多个国家，在沿海省份创造了近600个工作岗位（全职和季节性），为当地经济贡献了4000多万美元。ASL在新斯科舍省和新不伦瑞克省的岩栖草租约以及缅因州的收获区覆盖了3000多公里的海岸线，每年需要对这些海岸线进行调查，以获得适当的资源管理计划所需的必要信息。ASL目前花费了大量的时间和金钱进行实地调查，并且有兴趣研究分析卫星图像是否可以降低这一成本。ASL也有兴趣扩大他们收获的rockweed数量，并希望使用卫星数据来协助绘制资源分布地图，以确定潜在的新收获地点。位于新斯科舍省社区学院（NSCC）的应用地理信息学研究小组（agg）将与ASL合作，研究最新的高分辨率卫星Worldview-2在缅因湾低潮条件下获得的图像的利用情况。将采用数字图像处理技术，包括模式匹配和面向对象的分类方法来绘制岩杂草。

Acadian Seaplants Limited (ASL) is a globally recognized industry leader in the processing of seaweed-based products for food, biochemical, agricultural and agri-chemical markets worldwide and in the cultivation and processing of unique seaweeds for global food markets. The seaweed Ascophyllum nodosum (Rockweed) is the main raw material for the products produced. The end products are sold in more than 70 countries worldwide creating close to 600 jobs (full-time and seasonal) in the Maritime Provinces and contributing more than $40 M to the local economy. ASL's rockweed leases in Nova Scotia and New Brunswick along with harvested areas in Maine cover more than 3,000 km of shoreline that need to be surveyed on an annual basis to obtain the necessary information for proper resource management plan. ASL currently spends a significant amount of time and money conducting field surveys, and are interested in researching if analyzing satellite images can reduce this cost. ASL are also interested in expanding the amount of rockweed that they harvest and would like to use satellite data to assist with mapping the distribution of the resource to determine potential new harvest sites. The Applied Geomatics Research Group (AGRG) at the Nova Scotia Community College (NSCC) will collaborate with ASL to investigate the utilization of imagery from the latest high-resolution satellite, Worldview-2, acquired at low tide conditions at sites within the Gulf of Maine. Digital image processing techniques will be employed, including pattern matching and object orientated classification methods to map the Rockweed.