Knowledge based determination of plant species

Pollen for medical applications can be sorted by special machines. An integrated database and a spectrometric calibration allow a precise separation of the species with an identification of the plant species.

Short Description




Pollen is the principal food of the bees besides the nectar. Bees collect pollen in pockets on the back legs. Bee keepers gather the pollen lumps with a trap at the entrance of the bee hive. The pollen lumps are about 2-4 mm long and will be dried and sold in a mixed form.

This pollen is a valuable natural product with a multitude of applications in medicine and pharmacy but also as an adjunct for food. Especially in the allergy research pollen plays an important role, since pollen is held responsible for approximately 48% of all allergies.

In spite of this high potential the market for mixtures of pollen is very small, because of the difficulties with the investigation of the effects of pollen mixtures. To get new applications and markets it is necessary to sort the pollen according to the plant species.

In preceding research projects (Lebensmittelcluster OÖ, Fabrik der Zukunft 2nd call) a sorting machine was built and patented that was able to separate pollen according to the plant species with a very high purity and a flow rate of the approx. 1000 kg per year.

Project goals

Goal of this project was a knowledge based determination of the plant species of the sorting result additional to the separation of the species. This allows the producer of sorted pollen (e.g. consortium of bee keepers) a first determination of the plant species without costly laboratory tests.

To reach this goal the following questions had to be answered:

  1. Which data has to be collected in the sorting machine to allow a determination of the species.
  2. Which additional information about the pollen must be stored in a database to allow a determination of the species together with the measurements in the sorting machine.


Essential for the determination of the species is the calibration of the whole system due to the tolerances and factors of influence of the colour sensor. An identification of the spectrum of the illumination and a following correction of the colour space is therefore necessary in defined intervals. One of the results of the project is such a calibration method.

Furthermore, the factors of influence on the colour of the pollen have to be known. As a main factor of influence on the colour of the pollen, the date of gathering and the gathering region have been identified. Factors of influence concerning the drying and the storage of the pollen are of lower significance for the colour. Hence the origin and the date of gathering have to be documented and fed into the sorting machine to enable a unique species determination.

Hence a simple database was implemented, which provides the appropriate data during the runtime of the sorting machine to allow an identification of the plant species together with the colour values.

Project Partners

Project manager

Profactor Produktionsforschungs GmbH
DI Werner Reisner
Im Stadtgut A2
4407 Steyr-Gleink
Tel.: ++43 7252 885 0
Fax: ++43 7252 885 101

Project partners

  • Austria Consult Beratung und Organisation
  • IH-Tech Sondermaschinenbau und Instandhaltung GmbH
  • Lumitech Produktion und Entwicklung GmbH