作者: Cecilia Di Ruberto , Andrea Loddo , Lorenzo Putzu
DOI: 10.1007/978-3-319-68560-1_31
关键词:
摘要: In this work we propose a comparative study between different descriptors in analysing histological images. particular, our is focused on measuring the accuracy of moments (Hu, Legendre, Zernike), Local Binary Patterns and co-occurrence matrices classifying The experimentation has been conducted well known public datasets: HistologyDS, Pap-smear, Lymphoma, Liver Aging Female, Male, Gender AL CR. comparison results show that when combined with extracted from RGB images, orthogonal improve classification performance considerably, imposing themselves as very powerful for image analysis.