Analysis Of OSAHS Early Pathological Images Based On The Top-Hat Operator

Obstructive Sleep Apnea Hypopnea Syndrome(OSAHS) is a kind of oral disease with high prevalence rate and effecting one's health and sleeping seriously.Currently,doctors primarily observe the mouth with electronic laryngoscope instrument, the way to diagnose by eyes will not only increase the amounts of diagnosis, but also make a substantial increase the misdiagnosis rate.So this paper uses image processing methods to process and analyze OSAHS early pathological image, extract the edge of interesting region, then calculate the relevant medical parameters.This paper selects Top-Hat operators to process OSAHS early pathological image based on comparative analysis of several common mathematical morphology operators, and proposes multi-directional, multi-scale Top-Hat operators to improve it. Then introduces image acquisition and analysis, image preprocessing, contrast enhancement, edge segmentation, filling image, calculating the area and the diagnose processing of OSAHS disease successively. Finally, gets the conclusion that comprehensive Top-Hat operators have more advantages, can extract the edge of image more clearly, and have better edge closure through the processing of a large number of OSAHS early pathological image.This paper takes oral image, internal nasal passages image and throat at vocal image for example, uses multi-directional, multi-scale Top-Hat operators to enhance contrast of OSAHS early pathological images. Add the original gray image successively in contrast enhancement operations to avoid missing the useful information. Through repeated application of Top-Hat operations, the contrast of the interesting region and background is more and more big, can enhance edge information of interesting region, then use Morphological gradient to extract edge information. The processing results of images show that the operator proposed in thispaper can preserve image detail fully and make the edge information of the image more complete and accurate, edge closure of image is 97.67%. Lay a solid foundation for the later calculation of the relevant medical parameters of OSAHS early pathological image precisely, achieving electronic medical diagnose. This operators have good usability.