The Research Of Human Eyes Change Information Detection Technology

For many years, a number of studies found that there are more than70%of theinformation is gained from the human eye in human cognitive world and understand thesurrounding environment and the human communicate with each other to understandthe process. The human eyes may reflect a person’s mental state, psychological state,health state and so on. The researches of intelligent detection on the human eye changeinformation have become more and more the needs of all subjects, and promote therapid development of the subject. The human eye change information’s research has thegreat significance, regardless of the theory and practices.The main research of this thesis is that the human eye chance information carriesout detecting statistics. First on the face images of the face digital image preprocessingoperation, histogram modification technique for image enhancement, using gauss filterfor image processing, image sharpening, smoothing, and then the image gray scaleimage conversion processing; the human eye detection module achieved in two steps:using gray value difference compared to the human eye screening area, at first, removethe most of unimportant information in the face image. Then use template matchingalgorithm to detect the human eye position. The former algorithm is simple andcalculation is fast. The latter turns on the establishment of template, and templatematching algorithm and in very mature, choosing appropriate weight values will begood positioning the human eye and can remove a certain degree of the influence offace image background. Then the eye feature extraction, while calculating the width andheight, is used to introduce linear gray-scale image scanning approach for truncatedsegment length is the eye in the long axis of the stub shaft. And about the human eyeperimeter, area is improved method for the shape of calculated, the superiority of theproposed method in this paper through so many times test. Finally is the human eyechange information statistics module. The algorithm of this model is simple, as long asthrough different face image of the corresponding feature extraction eye comparison,through comparison results reflect on the change of the human eye information, andthen do the record. In summary，at the theory of all, development platform based on the Visual Studio2005, and by using C++on the visual change detection for programming. Experimentsshow that the algorithm of each module can be quickly and accurately achieve theexpected goal, and effectively the human eye to detect changes in information, withbetter practicability, reliability and scalability.