Research and development of a vision-based module for the classification of vehicles, which enables the using vehicle parts, ensures the recognition of more than more than 30 fine-granular classes. This to comply with all relevant international vehicle categorization systems.