An unsupervised framework for cytoplasm segmentation from partly superficial-overlapping cervical smear images
Cytoplasm segmentation is a subproblem of cervical smear image segmentation.Since the overlapping, complex morphology and other reasons, cytoplasm segmentation is a challenging problem.In this work, we propose an unsupervised framework for cytoplasm segmentation from partly superficial-overlapping cervical images.After denoising, the original images are divided into many superpixels using SLIC algorithm.Then 13 features of each superpixel are extracted and used to classify the superpixels to three classes by K-means++ clustering algorithm.Finally, we make postprocessing to deal with the miss-classification cytoplasm regions according to the histogram and other information.Experiments have been performed on two real-world partly superficial-overlapping cell image datasets, including the public Herlev dataset and one self-made cerivical smear image dataset show the effectiveness of the proposed method.The framework is promising and could be extended to further accurate segmentation one in the furure.
Cervical smear image Cytoplasm segmentation SLIC superpixel Superpixel feature extract Partly superficial overlapping
Lili Zhao Kuan Li Jianping Yin Jinzhao Yang
College of Computer Science,National University of Defense Technology,Changsha,410073 China Institute of Software,College of Computer Science,National University of Defense Technology,Changsha State Key Laboratory of High Performance Computing,National University of Defense Technology,Changsh PLA,000000 China
国内会议
金华
英文
1-7
2015-10-30（万方平台首次上网日期，不代表论文的发表时间）