• Argentina (español)
  • Austria (Deutsch)
  • Belgium (français)
  • Belgium (Nederlands)
  • Brazil (português)
  • Canada (English)
  • Czech Republic (čeština)
  • France (français)
  • Germany (Deutsch)
  • Global (English)
  • Great Britain (English)
  • Hungary (magyar)
  • Netherlands (Nederlands)
  • Poland (polski)
  • Ukraine (українська)
  • USA (English)
Digital Farming Products

Single Weed Identification

Single-Weed Identification

The algorithm has a 45% to 100% rate of accuracy in identifying more than 123 types of recorded weeds

Understanding how the algorithm works

Maintain coverage? In-Scope? Accuracy (%) Reliability
Early stages of growth ✔    moderate
123 weeds identified ✔  45%-100% high
Best 108 ✔  80%-100% high

Accuracy Range

88% of the 123 weeds are identified with an accuracy rate of over 80% (best 108). The accuracy rate can vary by growth stage. Sharp pictures which show an early stage of growth and individual weeds, have an accuracy rate that corresponds to the probabilities shown above.