Statistical Analysis of Cervical Overlapping Cells and Identification of Lesion Cells Based on Sparse Morphological Learning Model

Abstract

Clinically, the microscopic images of cervical cell smears are complex and changeable. Besides, the individual’s fluctuation is large, and the environment of dyeing is fluctuant, which causes the cervical cells in the smear microscopic images are of a complex category, and a large number of cellular agglomerations and adhesions when hyperplasia is active. Therefore, we proposed a statistical analysis of cervical overlapping cells and identification of lesion cells based on morphological learning model, which adopts sparse representation and mathematical morphological operation to achieve the segmentation of the complex cell from clinical smear microscopic images. The segmentation results demonstrate that the proposed method has the best segmentation performance and excellent generalization capability. It can be applied in the clinical application of cervical cytology images analysis. The proposed method has certain theoretical significance and application value.