|کد مقاله||کد نشریه||سال انتشار||مقاله انگلیسی||ترجمه فارسی||نسخه تمام متن|
|4539128||1626618||2016||10 صفحه PDF||سفارش دهید||دانلود رایگان|
• Object Based Image Analysis (OBIA) was used to map offshore Florida habitats.
• OBIA classification results were compared to traditional photo-interpreted maps.
• Accuracy of the OBIA results (78% overall accuracy) was greater than photo-interpretation results (71% overall accuracy).
• Thematic and spatial resolution of the OBIA map results were also higher.
• OBIA can provide a more quantitative, repeatable and faster alternative to traditional mapping methods.
The offshore extent of seagrass habitat along the West Florida (USA) coast represents an important corridor for inshore-offshore migration of economically important fish and shellfish. Surviving at the fringe of light requirements, offshore seagrass beds are sensitive to changes in water clarity. Beyond and intermingled with the offshore seagrass areas are large swaths of colonized hard bottom. These offshore habitats of the West Florida coast have lacked mapping efforts needed for status and trends monitoring. The objective of this study was to propose an object-based classification method for mapping offshore habitats and to compare results to traditional photo-interpreted maps. Benthic maps were created from WorldView-2 satellite imagery using an Object Based Image Analysis (OBIA) method and a visual photo-interpretation method. A logistic regression analysis identified depth and distance from shore as significant parameters for discriminating spectrally similar seagrass and colonized hard bottom features. Seagrass, colonized hard bottom and unconsolidated sediment (sand) were mapped with 78% overall accuracy using the OBIA method compared to 71% overall accuracy using the photo-interpretation method. This study suggests an alternative for mapping deeper, offshore habitats capable of producing higher thematic and spatial resolution maps compared to those created with the traditional photo-interpretation method.
Journal: Estuarine, Coastal and Shelf Science - Volume 181, 5 November 2016, Pages 83–92