The merge region algorithm is very helpful in:

• reducing the number of objects within a project
• reshaping your objects to finally receive your target objects
• preparing your results for vector export

The merge Region algorithm merges all neighboring image objects of a class into one large object. The class to be merged is specified in the domain. If you want to merge another class, you will have to create another process for that class. If you insert multiple classes into the Domain they are all merged together. If we would add the "grey" class into the Domain with the "blue" class (see example below) we would end up with only one object. Creating two single processes, one merge region for the "blue" and one merge region for the "grey" class we would end up with three objects and this is usually what you want to see.