Scan vehicle registration document and record data with AI | Update 2024

AI scanner software captures all data in seconds for the automotive industry, insurance companies and government agencies.

Optimize your processes with a vehicle registration document OCR system. Avoid the tedious manual recording of license plates, Vehicle identification number (VIN)HSN/TSN and other data on the registration certificate part 1 (formerly known as the vehicle registration document).

With an OCR scanner for vehicle registration documents, you not only save time, but also avoid errors and enable contactless customer processing in the age of Corona. Invest in a vehicle registration scanner for more efficient and reliable processing.

Anyone who has ever transcribed the data of a vehicle registration document knows the complexity of the document: license plate number, the vehicle identification number, or FIN for short in Germany, and the HSN/TSN are real error traps. On top of that, manual data entry costs a lot of time. Valuable minutes that are wasted with monotonous work.

Train digital vehicle registration scanners with AI

That's what some users of Konfuzio already thought through. Our users only need the high-quality software Konfuzio and 50 examples. Not the IT specialists tinkered, but even laymen train the AI, which can also work with other certificates, for example, for insurance companies, appraisers, offices or lawyers. As a result you get a digital vehicle registration scanner. "This saves up to 90 percent of the time previously spent on the project for the IT", says CEO Christopher Helm.

But how does the program work in practice? Quite simply, all that is needed is a photo or scan to upload, which can then be digitally read in seconds and converted into an editable format. Thanks to the use of artificial intelligence, the software also does not fail to recognize the forgery-proof document like ordinary text recognition does.

Digital vehicle registration OCR with smart AI

While many readout applications use so-called OCR engines (optical character recognition), users of Konfuzio are helped by the artificial intelligence (AI) developed in-house. This makes the vehicle registration document scanner "adaptive", and the performance of the software improves over time.

The solution processes data reliably even under adverse conditions - for example, with photos that are poorly exposed or were taken from unfavorable angles. Thanks to the integrated plausibility check, the system can even fill in missing sections of an incomplete document.

How does a digital vehicle registration document scanner work?

  1. A photo or scan of the registration certificate will be taken.

  2. The document is uploaded via the API (interface).

  3. The AI software processes the document.

  4. The extracted sections of the ZBI are transferred and integrated in JSON format.

With its certified security structure, Konfuzio offers the highest standards in data protection and DSGVO compliance. The vehicle registration document OCR can also be installed on-prem or in your private cloud. You can find more information in our article on the topic of AI Cloud. Über die API-Schnittstelle kann die Software in eine vorhandene IT-Infrastruktur integriert werden.

Which fields can the vehicle registration document OCR recognize?

  • B: Date of first registration
  • D.1: Brand
  • D.2: Type/variant/version
  • D.3: Trade name(s)
  • E: Vehicle identification number
  • F.1: Technically permissible maximum mass in kg
  • F.2: Total permissible mass in kg in the member state of registration
  • G: Mass of the vehicle in service in kg (unladen mass).
  • H: Validity period
  • I: Date of this approval
  • J: Vehicle class
  • K: Number of the EC type-approval/ABE
  • L: number of axes
  • O.1: Technically permissible trailer load with brakes in kg
  • O.2: Technically permissible trailer load, unbraked, in kg
  • P.1: Displacement in cm³
  • P.2/P.4: Rated power in kW/nominal speed at min¹
  • P.3: Fuel type or energy source
  • Q: Power-to-weight ratio in kW/kg (only for motorcycles)
  • R: color of the vehicle
  • S.1: Seats including driver's seat
  • P.2: Standing room
  • T: Maximum speed in km/h
  • U.1: Stationary noise in dB (A)
  • U.2: Speed in m (high -1) to U.1
  • U.3: Driving noise in dB (A)
  • V.7: CO₂ (in g/km) combined value
  • V.9: Pollutant class relevant for EC type-approval
  • (2): Manufacturer short name
  • (2.1): Code to (2)
  • (2.2): Code for D.2 with check digit
  • (3): Check digit for vehicle identification number
  • (4): Type of structure
  • (5): Designation vehicle class/bodywork
  • (6): Date to K
  • (7): Technically permissible maximum axle load/mass per axle group in kg
  • (7.1) to (7.3): Axis 1 to axis 3
  • (8): Permissible maximum axle load in the member state of registration in kg
  • (8.1) to (8.3): Axis 1 to axis 3
  • (9): Number of drive axes
  • (10): Code to P.3
  • (11): Code for R
  • (12): Volume of the tank in tank trucks in m³.
  • (13): Support load in kg
  • (14): Designation of the national emission class
  • (14.1): Code to V.9 or (14)
  • (15): Tire
  • (15.1) to (15.3): on axis 1 to axis 3
  • (16): No. of the registration certificate part II
  • (17): Operating permit characteristic
  • (18): length in mm
  • (19): Width in mm without mirrors and attachments
  • (20): Height in mm

Learn more about Konfuzio now

Konfuzio offers various advantages over other providers

AI model for vehicle tickets on Konfuzio marketplace

Are you looking for a solution that efficiently and accurately extracts data from the German vehicle registration document? The pre-trained AI model for vehicle registration documents from Konfuzio has the right solution for you on the marketplace.

KI model vehicle registration document DE sample Konfuzio marketplace

Register for free at and see for yourself. Do you have any questions or need support? Write us a message.

Florian Zyprian Avatar

Latest articles