Study On Detection Method Of Abnormal Image Content In Liquid Based Of Cell

In this paper,liquid based cervical cell producer abnormal cell image as the research object,described and analyzed by the physical characteristic of the abnormal image of liquid based cervical cell producer,establish abnormal attribute model of liquid based cervical cells,the use of statistical methods for liquid medium cervical cytology abnormal specimens of different physical parameters were recorded,and the combination of multi sectional set segmentation method to obtain the liquid based cervical cell producer abnormal cell image feature parameters to realize the automatic recognition and detection of liquid based cervical cytology abnormal quality.The specific research contents are as follows:(1)combination of knowledge and human visual perception of medical pathology,from visual perception of visual consistency of abnormal image of liquid based cervical cytology,and verified by the experiment of abnormal image of liquid based cervical cell producer color,intensity and area change curve feature consistency.(2)according to the liquid based cervical cell producer abnormal image feature consistency principle,the establishment of liquid based cervical cell producer abnormal image cell physical property model,and verified by the experiment of visual features of normal liquid based cervical cytology abnormal images and liquid based cervical cell producer distinction,provides the data basis for detection of abnormal image liquid based on the production of cervical cells.(3)the visual characteristics of liquid based cervical cell producer anomaly analysis,put forward a multi threshold image segmentation method set,establishing multi sectional set image segmentation model,and through the experiment to verify the accuracy of multi sectional set segmentation method,and zero parameter values of the validity and reliability for abnormal image liquid based cervical cell producer provides decisive parameters,realize the parameter identification and detection of liquid based cervical cytology abnormal cell image.(4)based on rough set theory,the characteristic parameters of liquid based cervical cytology abnormal cell image,according to rough set model,the object with the form set will be studied are described and classified,described the establishment for a collection of object attributes,the conditions set by study to distinguish,to achieve automatic recognition with the detection of liquid based cervical cytology abnormal cell image.Experimental results show that the set of the rough set model and multi sectional set segmentation model can effectively realize automatic identification and detection of abnormal image of liquid based cervical cell producer,on cervical cell image processing analysis before the unqualified specimens were removed,provides a watchdog mode for cervical samples quality control.