Triton Sediment - Manual (Page 9)

Page 6
2.2 Create Training Set
The first step in the classification process is to enter ground truth information to
generate a set of data points to be used in the neural net training process. This can
originate from bottom sampling programs or video/diver surveys, or can be entered
manually if no bottom information is available.
To create the training set, click on the toolbar button or right-click on the Manual
Training node in the SeaClass File Tree and select 'Create' as shown below:
This will open the 'Create Training Set' window shown below:
Options available on this page
include:
Training Set Name - Allows the
user to give the training set a
unique name.
Sampling Grid Size - Sets the
output resolution in meters. If
'Custom' is selected, the value
is entered in the 'Custom Grid
Size' box below.
******************************
Important Grid Size Note!!!
Selecting the correct grid size is likely the most important factor for getting great
results! Therefore it is very important for the user to understand how this relates to the
final results.
Perspective SeaClass uses a textural segmentation algorithm for determining bottom
types in an image (see Appendix A for more information on how SeaClass works!). During