can’t remember, sorry. Probably the default value.
regarding the logic, please have a look at the documentation. The pixel size times the window size is the extent of your background window in meters. It should be large enough to cover single oil spills.
Adaptive Threshold Algorithm
The dark spots are detected using an adaptive thresholding method.
- First the local mean backscatter level is estimated using pixels in a large window.
- Then the detecting threshold is set k decibel below the estimated local mean backscatter level. Pixels within the window with values lower than the threshold are detected as dark spot. k is a user selected parameter (see parameter Threshold Shift below).
- Shift the window to next window position and repeat step 1 and 2.
Discrimination
- First the contiguous detected pixels are clustered into a single cluster.
- Then clusters with their sizes smaller than user selected Minimum Cluster Size are eliminated.
Visualize Detected Oil Spill
The oil spill detection bit mask is output as a separated band. To view the oil spill detection results, the following steps should be followed:
- Bring up the image.
- Go to Layer Manager and add the oil spill bit mask band as a layer.
Parameters Used
For dark spot detection, the following parameters are used (see figure 1):
- Source Bands: All bands (real or virtual) of the source product. User can select one or more bands for producing multi-looked images. If no bands are selected, then by default all bands are selected.
- Background Window Size: The window size in pixels for computing local mean backscatter level.
- Threshold Shift (dB): The detecting threshold is lower than the local mean backscatter level by this amount.