Extraction Of Iris Image Features And Medical Applications

The iris diagnosis theory asserts that the changes of the human iris feature reflect the level of human health.With the development of technology,the medical auxiliary diagnosis system becomes more popularized and applied,so the development of the iris auxiliary diagnostic system will provide powerful technical support for iris diagnosis.The curling wheels,the holes,the pigment spots,the radiation gutters and their locations are the representation of human iris features,the iris auxiliary diagnosis system first realizes the extraction of the above iris features,Digital image processing technology is used in this paper to complete iris localization and iris features extraction.And on this basis,iris auxiliary diagnosis system is built combined with the image acquisition equipment and the iris diagnosis theory.This paper mainly completes the following work:1.Image preprocessing.According to the characteristics of eye pupil,the method combined image inverse complement with PSO is used to complete the fitting of the iris inner circle.The method combined zoning domain denoising with PSO which can overcome the light spot and complete the segmentation of the iris region effectively is used to realize iris outer circle localization on the basis of the characteristics of the iris outer boundary.Then the method of coordinate change is used to normalize the image after localization.2.Extraction of curling wheel.According to the characteristics of the rolling wheel,the method of structural pattern recognition is used to extract the curlingwheel,which can effectively avoid the interference of radiation trenches and eyelashes to depict outline of a curling wheel.3.Extraction of pigment spots.According to the gray scale of pigment spot,the method of K-S detection algorithm,which converts the problem that extracting the pigment spot into the problem that looking for the probability distribution of each representative in the image,is applied to accomplish extraction of pigment spots.This method can cover pigment spots well and reduce the original morphological characteristics of pigment spots.4.Extraction of holes and radiation gutters.According to the morphological characteristics of the hole and the radiation gutters,firstly,all the target textures are extracted by the method of physique,and then the SVM classifier is applied to classify the holes and radiation gutters from the interference features,at last,setting the determinant condition of radiation gutters to determine the radiation gutters,above all,the extraction of the radiation trenches and holes is realized.This algorithm can effectively restore the morphological characteristics of features while the feature extraction is realized.5.Roughness calculation.the concept of roughness,one of the Tamura texture features,is used to calculate the roughness in iris represent area,and based on the concept of iris density,the feasibility analysis that replacing density with roughness is completed.6.The establishment of iris auxiliary diagnosis system.Combined with the extraction of iris features,the rules of iris feature description and pathological are established on the basis of the iris diagnostic rules,and the iris auxiliary diagnosis system is programmed.The system can be applied to diagnose subhealth of human body and organ chronic diseases on the basis of the iris diagnosis theory.