Cervical cell segmentationYouyiSong et.al (2017) proposed a frame work of deep learning technique anddeformation model to segment cervical cell accurately from overlapping clumps.This frame work not only helps in shape correlation of overlapping cells butalso by the proposed method it is helpful in utilizing the cell structurecontextual information, the proposed convolutional neural network (CNN) learnsfeatures at different scales. Therefore, contexts of multiple scales areintegrated into local decisions to correctly classify each pixel intobackground, cytoplasm, or nucleus pixel.Multiplecells labeling method is proposed to get accurate splitting of detectedoverlapping cytoplasm.
The classi?ed cytoplasm pixels are labeled with thecorresponding nucleus labels generated by CNN. Gaussian kernels ?tting is usedto estimate the shape of the cell, therefore the dependence of color informationis obtained by the shape cues. Two different data sets have beenused in the proposed framework and the segmentation results are compared withthe state-of-the-art approaches. Additionally, theproposed method is also effective in segmenting abnormal cells, even for imageshaving large number of overlapped cells and high degrees of overlapping.ZhiyunXue et.al (2010) analyzed and separated clinically a definite area from uterinecervix image in a huge database created for the research of cervical cancer.Boundary Marketing tool segmentation (BMT), cervigram segmentation tool (CST),Multi-observer segmentation Evaluation system (MOSES) are the method followedin medical image processing.
Cervicographic image processing algorithm is used todesign for Cervicographic image processing. The proposed algorithm is nominalin cost in late years. Similar to colposcopy, Cervicographic is based on theaceto whitening, a transient phenomenon observed when a diluted acetic acidsolution is applied to the surface of the cervix.