In many biological studies, the species identification of the organisms is a crucial element, which may not be easyeven for experienced biologists. New technologies thus represent ideal tools for both simplification andacceleration of identification of organisms. In recent years, a lot of attention has been paid to mobile softwareapplications, which can combine scientific disciplines such as mathematics, informatics and biology. Suchsoftware can work based on available species occurrence records in combination with their ecological andbiogeographical characteristics, as well as photographs and distinctive sounds. In our case, we have focused onimages, commonly taken by photographers or biological enthusiasts for many organisms and thus an easilyaccessible and usable source of information for professionals. Pilot research conducted on freely available onlineimages of dragonflies and damselflies has shown promising results on both a mobile device-sized model and astandard, computationally more complex, model. We have achieved ~54%/~74% top-1/top-5 testing accuracy onthe standard model and ~47%/~70% top-1/top-5 testing accuracy on the mobile model. This method foridentifying species from the image is a very promising way how to make from an ordinary observer, when usingappropriate tools, an ?experienced? field biologist. The accuracy of the combination of a system using occurrencedata with the species characteristics and recognition of image material can then be a breakthrough approach forspecies identification