Comparison of SURF and HOG extraction in classifying the blood image of malaria parasites using SVM
| Autor: | Muhathir, Rollys Gultom, Julianus Stepanus Sihotang, Reyhan Achmad Rizal |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
010302 applied physics
biology business.industry Computer science Extraction (chemistry) Feature extraction Anopheles 020206 networking & telecommunications Pattern recognition 02 engineering and technology medicine.disease biology.organism_classification 01 natural sciences Support vector machine parasitic diseases 0103 physical sciences 0202 electrical engineering electronic engineering information engineering medicine Artificial intelligence business Malaria Plasmodium species |
| Zdroj: | 2019 International Conference of Computer Science and Information Technology (ICoSNIKOM). |
| DOI: | 10.1109/icosnikom48755.2019.9111647 |
| Popis: | Malaria is the deadliest disease caused by a protozoa parasite of the Plasmodium genus that is transmitted through the female Anopheles mosquito. Five Plasmodium species can cause infections and deaths such as P. Ovale, P. Malaria, P. vivax, and P. falciparum, each parasitic malaria pattern in the blood image has a unique and distinct pattern. The introduction of the parasitized pattern and uninfected cells conducted in this study used the extraction of surf and HOG features as well as SVM as methods classification. Research results show that SVM with SURF feature extraction can classify the parasitized pattern and uninfected cells in the blood image better than SVM with the extraction of HOG features. The average accuracy feature extraction of the SURF 85.83% while HOG feature extraction is only 83.50%. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|