Document separation
Document separation is executed once per batch, and on the document level. If the batch contains loose pages, these are combined into a single document. If more than one document is available, they can be processed in parallel by Server. Due to the fact that document separation can be parallelized, no XFolder data, such as xvalues (pXRootFolder.XValues), can be modified during separation.
Before separation, the pages where a document is split are marked. It is possible to use scripting to modify which pages are marked for separation. After all information is available, the server executes the separation.
By default, document separation is not enabled. You can activate document separation in the project settings. You have the choice between two different separation algorithms.
Trainable Document Separation (TDS)
Separating documents by the standard document separation the classification result of the single pages is used to recognize the first page of a document. TDS is using enhanced algorithms for this separation process, please refer to the Project Builder Help for details.