The Stury Of Iris Frill Detection Methods Of Iris Diagnosis

Iris diagnostics is also called iridology, it is a method to confirm internal potentialpitfalls, pathological changes and functional disorder of human body’s organ throughcheck iris features. Iris diagnostics is not the diagnosis of disease, but the determination ofhuman body’s each organ, organization and system function and weak degree throughobserving the iris abnormal condition. So the technology is helpful to find out the potentialdisease and health risk earlier, and make timely diagnosis and treatment to help peoplemaintain or recover healthy condition. In the iris region partition, iris frill is the boundaryof the intestinal and visceral area, and at the same time, the position of iris frill can displaythe intestinal health.Based on the analysis of the present research, the research of iris frill extraction is stillin its initial stage, there is no perfect method theory. Aiming at the problems in presentresearch, this paper puts forward a detection algorithm based on image gradient extreme,this algorithm can realize the iris frill detection automatically.This method uses the gray difference between the iris frill and its surrounding areas.First, the color image should be turn into a gray image, and using the thresholdsegmentation method to extract the pupil from the gray image. Second, according to thepupil’s information the iris position which including the iris frill parts will be located, andthen normalizing the ring iris image into a rectangle one. Thirdly use the improvedgradient operator to do the differential operation in the rectangular image, then the negativepoint of the extreme value will be found in the definition of the range query, and thesecoordinates are the iris frill’s profile points. Finally restore the contour points to theoriginal gray image to finish the extraction.In this paper, MATLAB was used to design and compile the algorithm. To the imagewith a relatively clear iris frill’s profile, this algorithm had a good effect. To the laboratoryiris database, the location accuracy of this algorithm can reach92.5%. Experimental resultsshowed that this algorithm had many advantages in the feasibility, complicated degree andextraction efficiency. Through this study, it provides a technical basis for the automatic extracting, recognizing, matching and analysis of the iris anomaly characteristics in thesubsequent research.