Research On Image Acquisition And Image Processing Of Infant’s Fingerprint

Objective: To explore the application of image acquisition and processing technology inthe traditional Chinese medicine fingerprint of infants. To realize objective, digital studyof fingerprint image through research on the design of image acquisition hardware,clinical fingerprint image acquisition and fingerprint image processing and analysis, so asto explore the theory of Chinese medicine and the characteristics of treatment method.The thesis contains5chapters:1Introduction;2Infant’s fingerprint acquisition hardwaredesign;3Infant’s fingerprint clinical sample collection;4Infant’s fingerprint imageprocessing;5Conclusion and discussion.Methods: Using Microvision colorful digital camera of high resolution combined withmanual zoom lens and an annular LED visual light source to build image acquisitionhardware equipment and image acquisition software platform. Acquisition of infant’sfingerprint image with different disease, gender, age and record the clinical symptoms,signs and the changes of state of illness information. Study on the current imageprocessing, segmentation method, and analyzing the infant’s fingerprint image usingMATLAB software platform, comparing the effect of image processing.Results: The image acquisition system can objectively image, real-time observation offingerprints, fingerprint image acquired simultaneously can be preserved for a long time,to facilitate later processing observation and analysis. Various image processingalgorithms can clear tell infant’s fingerprint off position and length of characteristicfingerprint. After image processing, the results shows that it agrees with the naked eye observation of fingerprint features. Through the discussion of each method, experimentswere performed using the MATLAB software, we figure out the advantages anddisadvantages of the image segmentation algorithms. Image edge detection method hasdistinguished effect for extraction of fingerprint features, and Canny image segmentationalgorithm is superior to the other algorithms. Level set algorithm has distinguished effecton analysis of the index finger contour and length of fingerprint. Three-dimensionalfingerprint image can objectively shows the fingerprint features.Conclusion: This paper has accomplished using color digital image camera to captureclinical pediatric fingerprint image, which is objective, accurate and convenient. At thesame time, using edge detection, image segmentation based on MATLAB to get theregion of interest, so as to realize the objective collection and study of infant’s fingerprinton TCM, and avoid errors of clinicians’ subjective diagnosis, and then provides areference for clinical diagnosis.