Publicação
Pollen grain recognition through deep learning convolutional neural networks
| Resumo: | Palynology is the study of pollen, in particular, the pollen’s grain type, but the tasks of classification and counting of pollen grains are highly skilled and laborious. Despite the efforts made during the last decades, the manual classification process is still predominant. One of the reasons for that is the small number of taxa usually used in previous approaches. In this paper, we propose a new method to automatically classify pollen grains using a state-of-the-art deep learning technique applied to the recently published POLEN73S image dataset. Our proposal manages to classify up to 94% of the samples from the dataset with 73 different classes of pollen grains. This result, which surpasses all previous attempts in number and difficulty of taxa under consideration, gives good perspectives to achieve a perfect score in pollen recognition task even with a large number of pollen grain types. |
|---|---|
| Autores principais: | Monteiro, Fernando C. |
| Assunto: | Pollen recognition Deep learning Convolutional neural network |
| Ano: | 2022 |
| País: | Portugal |
| Tipo de documento: | comunicação em conferência |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Instituto Politécnico de Bragança |
| Idioma: | inglês |
| Origem: | Biblioteca Digital do IPB |
| Resumo: | Palynology is the study of pollen, in particular, the pollen’s grain type, but the tasks of classification and counting of pollen grains are highly skilled and laborious. Despite the efforts made during the last decades, the manual classification process is still predominant. One of the reasons for that is the small number of taxa usually used in previous approaches. In this paper, we propose a new method to automatically classify pollen grains using a state-of-the-art deep learning technique applied to the recently published POLEN73S image dataset. Our proposal manages to classify up to 94% of the samples from the dataset with 73 different classes of pollen grains. This result, which surpasses all previous attempts in number and difficulty of taxa under consideration, gives good perspectives to achieve a perfect score in pollen recognition task even with a large number of pollen grain types. |
|---|