The sensors of new-generation commercial satellites and the image processing techniques currently make possible to study and monitor shoreline variations. One of the main problems lies in locating the shoreline via a method that is not bound to the “subjectiveness” of the interpreter. A method and a software-based procedure for measuring the water/sand interface was developed to measure the position of the shoreline, based on the diff erent spectroradiometric response of water and sand in the red and infrared wavelengths. The vegetation index (NDVI) computed as the ratio between infrared and red bands has indeed proved to be particularly suitable for the identifi cation of the water/sand interface. This method is also applicable when analysing images acquired in diff erent periods, as required by monitoring programmes. The method was tried out in two diff erent sample areas along the Latium coast and was compared with traditional measuring techniques (photo interpretation), ground survey data (GPS measurements) and laboratory data (spectral signatures) that confi rm that NDVI values may be successfully used to identify the water/sand interface from satellite images taken with diff erent sensors. The procedure was tested on Quickbird and Ikonos images, using the software ENVI
A new approach to detect shoreline from satellite images.
SALVATORE, MARIA CRISTINA;
2008-01-01
Abstract
The sensors of new-generation commercial satellites and the image processing techniques currently make possible to study and monitor shoreline variations. One of the main problems lies in locating the shoreline via a method that is not bound to the “subjectiveness” of the interpreter. A method and a software-based procedure for measuring the water/sand interface was developed to measure the position of the shoreline, based on the diff erent spectroradiometric response of water and sand in the red and infrared wavelengths. The vegetation index (NDVI) computed as the ratio between infrared and red bands has indeed proved to be particularly suitable for the identifi cation of the water/sand interface. This method is also applicable when analysing images acquired in diff erent periods, as required by monitoring programmes. The method was tried out in two diff erent sample areas along the Latium coast and was compared with traditional measuring techniques (photo interpretation), ground survey data (GPS measurements) and laboratory data (spectral signatures) that confi rm that NDVI values may be successfully used to identify the water/sand interface from satellite images taken with diff erent sensors. The procedure was tested on Quickbird and Ikonos images, using the software ENVII documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
