Preliminary study of Mustelus schmitti's ova morphometry using Digital Image Processing techniques

CLAUDIO DELRIEUX; ELIAS FERNANDA; GISELA NOELIA REVOLLO SARMIENTO; NATALIA V. REVOLLO SARMIENTO

Conferencia; Sharks International Conference; 2018

Mustelus schmitti distributes geographically from the north of the Rio de Janeiro state (Brazil) to the south of Argentinian Patagonia, and shows only a left ovary. This chondrichthyan fish constitutes one of the main resources for fisheries in Argentina although captures are dropping substantially because overexploitation exerted a negative impact on their reproductive cycle. In this work we develop specific image processing techniques to automatically estimate the number, shape and dimensions of ova from the mustelus schmitti ovaries? microphotographs. The technique is based on a color-based tissue segmentation in dyed histological sections. Samples were photographed using a C7070 Olympus camera adapted to an Olympus BX51 light microscope. Images were converted to YIQ color space, which further facilitated a clustering operation in color space. After clustering, ova can be automatically segmented and their shape and size parameters can be precisely estimated. Preliminary results show more numerous albeit smaller ova in neonate and young specimens than in mature ones, which represents accurately the natural process of folliculogenesis.