Create Simple Pipeline

[FHorschig]

The pipeline should take an image and a annotation-helper as input and provide a method to select which algorithm(s) should be executed on that image.

[FHorschig]

Decouple annotator and analyzer.

[FHorschig]

Should the annotator really be decoupled from the analyzer? This would mean to store a lot information about single bugs like species, REAL coordinates (regarding that they are probably found on a cropped and scaled image) and image URI.