Handwriting Verifier with Help of Combined SVM-HMM Classifier Used with Curvelet Transformation

Handwriting Verifier is considered as important research field in the filled of forensic and biometric applications. It finds significance in fields like graphically which exploit the physiological performance of the human based on the handwriting. At this time too many technique are available for Handwriting verifier. Although no one of the techniques is yet proved to be clarify for large number of object. That's also fact that all the pattern of writing will be differ of any human with the time. HMM is the best technique for the verifier the writing for large number of object but its vector feature give differ patter verifier like retina verifier, used their training and test sample may vary. Hence Verifier of same tough. Therefore in this work we propose a technique for Handwriting Verifier with help of combined SVM and HMM. In this work curvelet transform are used predominantly for alphabet and numeric verifier problem and hence are more suitable for this work. SVM is also given good efficiency but not in the large object. Hence we develop a new classifier and show that the method performs better than self-sufficient HMM and SVM classifier.