Résumé
Accurate knowledge of the identity of plants underpins the success of agricultural development and biodiversity sustainable management and conservation. Unfortunately, such data are often hardly available for professional stakeholders, as well as for teachers, scientists and citizens. Furthermore, these informations are often incomplete for ecosystems with the highest plant diversity, i.e. Mediterranean and tropical regions. It is thus of the utmost importance to speed up the accumulation of basic data on plants, while simultaneously providing an easy and efficient access to this knowledge to users.The Pl@ntNet project will contribute to this objective by providing, within a coherent platform, free, open-source, easy-access software tools and methods for plant identification. It will involve citizen science, as a powerful means to enrich databases with information on plant description. These tools, partly based on well established, or prototype software, will be proposed for individual or collaborative use, to all potential users. The plant identification systems will include not only existing, interactive morphological recognition systems, but also more experimental, yet potentially powerful content-based image retrieval methods. Preliminary evaluations have been done on species identification by wood anatomy, and leaves morphology