Image Segmentation for the Diagnosis of Problems in the Manufacture of Equipments

This project proposes to study and develop segmentation methods based on evolutionary interval fuzzy algorithms and to evaluate them in comparison with other methods of segmentation of the three classes above (threshold, border-based and region-based). Specific objectives of this project are: i) To study models of fuzzy cellular automata and fuzzy genetic algorithms and their applicability in image processing; Ii) to develop image segmentation models based on the methods of item i), adding the characteristics provided by interval mathematics and type-2 fuzzy logic for the treatment of the uncertainty inherent in this type of problem; And iii) produce a separate evaluation of the methods when applied to different types of equipment components.

Publications:

- WUW SIBGRAPI 2014: Detecção Automática de Defeitos na Fabricação de Equipos Utilizando Processamento de Imagens

- Undergraduate Final Work (UFPEL): Estudo e Aplicação de Métodos de Segmentação de Imagens para o Diagnóstico de Falhas na Fabricação de Equipos