Estudo e implementação de técnicas para identificação de anuros em imagem

This paper describes the process used in extracting characteristics of Anurans in pictures and it makes a comparison between J48, Random Tree and Random Forest classification algorithms. Two scenarios for evaluation were defined: in the first one we used two classes to represent objects in the image...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Queiroz, Angélica Larissa M., Silva, César Alberto da, Jubileu, Andrea Padovan, Ruiz, Linnyer Beatrys
Formato: Objeto de conferencia
Lenguaje:Portugués
Publicado: 2016
Materias:
Ciencias Informáticas
Segmentation
Patterns
recognition
frog
Acceso en línea:http://sedici.unlp.edu.ar/handle/10915/55787
Aporte de:
SEDICI (UNLP) de Universidad Nacional de La Plata
Descripción
Sumario:This paper describes the process used in extracting characteristics of Anurans in pictures and it makes a comparison between J48, Random Tree and Random Forest classification algorithms. Two scenarios for evaluation were defined: in the first one we used two classes to represent objects in the image and in the second scenario five classes were used. The Random Forest algorithm obtained best result in the classification of data in relation to J48 algorithms and Random Tree.

Ejemplares similares