Research On Key Technology Of Fingerprint Recognition Based On Mobile Platform

Because the fingerprint of the unique,life-long invariance and good security,making it the world's most common application of the various areas of information.However,there are still some problems before and after the traditional fingerprint image acquisition.Such as the ridge line before and after the collection,fingerprint image distortion,mining area is small and health and other issues,these problems are difficult to get a better solution.And touchless fingerprint recognition technology not only can solve these problems very well,and it has a better recognition rate,stability and reliability.So it has become a good future research prospects.Today's mobile platform,the largest proportion of smart phones almost all with a camera,which for our mobile platform to develop touchless fingerprint recognition technology provides a natural advantage for future popularity to reduce development costs.The most widely used in the mobile platform is the Android system.Android is not only free,but also open source.Therefore,we will be in the Andrews mobile phone system to develop non-contact fingerprint recognition related technology.First of all,the characteristics of the non-contact fingerprint image acquisition mode,as well as the use of camera fingerprint image acquisition may bring the fingerprint image rotation,translation,zoom.We use a screen in the phone to draw the corresponding finger area contours of the method,the distance between the finger and the camera fixed,effectively overcome these problems.In the touchless fingerprint image preprocessing,the segmentation of the fingertip part is completed based on the YCbCr model and the OTSU algorithm,and the region of interest is determined by the fitting method of the finger locating circle.The resulting image is then preprocessed and displayed on the handset.Then,with the actual situation,try to use the SIFT algorithm to extract and match the touchless fingerprint image.In the course of the experiment,we found that there were more misplaced points.So we use the RANSAC algorithm,use it to delete the corresponding mismatched match pairs to get better results.In order to improve the speed of the system,the use of FLANN-based matching device.Experiments show that from the recognition accuracy,computing speed,storage space,etc.can be seen to meet the real-time requirements.Finally,this article uses Android development tools and OpenCV to achieve the above theory.And then analyzed the design of the subsystems of the work flow and the function of each module.Experiments show that the proposed acquisition module can effectively solve the problem of rotation and translation of touchless fingerprint image acquisition,and the designed subsystem can run stably.