Pap Smear Images Segmentation for Automatic Detection of Cervical Cancer

Ayubu Hassan Mbaga, Pei ZhiJun

Abstract: This work focuses on segmentation of Pap smear cervical cell image for automatic screening and detection of cervical cancer at early stage. This paper proposed linear contrast enhancement and median filter for removing of noise, sharpens and preserving edges and boundary of cytoplasm and nucleus. Canny detector algorithm was preferred and applied to a cervical cell images with the value of sensitivity of 0.634 and value of sigma was 6.56. We obtained the gradient images with smooth edges and boundary of cytoplasm and nucleus. Otsus algorithm was used to separate cytoplasm from the background. Maximum gray gradient difference (MGLGD) method adopted to extract nucleus contour. The results shows that segmentation gives impressive performance which will help further steps of automatic screening and detection of cervical cancer from Pap smear cervical cells image.

Keywords: Cervical cancer, Segmentation, Canny, MGLGD

How to Cite?: Ayubu Hassan Mbaga, Pei ZhiJun, "Pap Smear Images Segmentation for Automatic Detection of Cervical Cancer", Volume 4 Issue 4, April 2015, International Journal of Science and Research (IJSR), Pages: 940-943, https://www.ijsr.net/getabstract.php?paperid=SUB153074, DOI: https://dx.doi.org/10.21275/SUB153074