|
Acadêmico(a): Arno Wilson Cassaniga |
Título: PLANTARUM - API para Reconhecimento de Plantas |
|
Abstract: |
This work presents an Application Programming Interface (API) prototype for plant classification based on leaf images processing. By extracting geometric features, color features, venation, texture and morphological descriptors, the designed API is able to infer the plant specimen a given leaf belongs to, by comparing this feature set with a specimen database previously populated. The results show the designed API is efficient regarding plant classification, however adjustments are still required to remove the need of human intervention locating the petiole and leaf tip.
|
|
|