Fuzzy Segmentation of Cervical Cytology Image using Level set algorithm

Due to the low resolution and low contrast, image segmentation in medical profession is a very tough task. Due to their substantial reliance on human inputs, the widely used conventional segmentation approaches such as threshold, fuzzy-c-means have some drawbacks. These techniques are not able to capture the inherent uncertainties in an image. When the photos are distorted by noise, outlier and other artifacts, the performance further declines. In order to overcome the abovementioned problem, a new technique is proposed in this paper. This paper proposes a method for cervical cytology image segmentation based on Fuzzy c-means and the Level set algorithm. The current study has begun its work with a median filter with a fuzzy level set method to segment cervical cytology images. An image of cervical cytology was used as an input image. A median filter and fuzzy c-means were applied to the image to abstract image noise and generate image clusters, respectively. The image clusters displayed the initial and final cluster centres. Following separation and extraction of white matter from grey scale images, the proposed level set method was used for segmentation of cervical images. The fuzzy c-means were sensitive to the initial cluster centre. A new fuzzy level set algorithm can directly evolve from the initial segmentation by spatial fuzzy clustering. The results are also used to estimate the level set algorithm's controlling parameters. Additionally, the fuzzy level set algorithm benefits from locally regulated evolution. Such enhancements make level set manipulation easier and lead to more reliable segmentation. When compared to techniques described in prior studies, this approach fared well in segmenting cervical pictures The accuracy of the proposed method is 94%