The Application And Research Of The Image Processing Technology In The Development Of Handwriting Identification System

Nowadays, the signature has become a widely accepted the method as personal identification in the world. The study of automatic signature verification system based on the different data acquisition ways can be classified into 2 types: Off-Line signature verification, On-line signature verification. So far, more research for the On-Line signature verification, compared with Off-Line signature verification is also more mature. At present automatic signature verification instrument generally to use in the on-line appraisal in the developed country. Simultaneously the Off-Line aspect research also develops widely. Although the Off-Line aspect research has also made the certain progress, it still was an extremely difficult question in the handwriting appraisal domain, at present still could not achieve the practical stage.The dynamic feature of Signs has the appraisal significance has been appraised by the experts to recognize, however in the Off-Line signature verification system lacks is precisely the dynamic feature information. After research China Criminal Police University's signature verification expert discover that between the signature static feature information and the dynamic feature information is the existence certain relation. If found the rule between them through the thorough research, Off-Line signature verification System, adopts by the static feature information deducing dynamic feature information means, may obtain more significant the signature feature. That will enable the Off-Line signature verification system the performance to obtain the remarkable promotion.In order to discover the rule between the dynamic information and the static information exists, it needs carrying on the dynamic information and the corresponding static information the connection, in this process a question appeared that is matching dynamic information gained through the tablet and the signature scanning picture. This article through to the correlation domain analysis research, has carried on the experimental analysis to the existing picture matching technology, and aimed at the tablet particularity in this topic to propose the simple suitable matching method, completed the preparatory work for sought the rule between signature dynamic information and static information; This topic also aims at the signature handwriting dynamic information sample the particularity, proposed one kind of stroke extraction algorithm, facilitates to extracts the width of stroke, in the foundation of the stroke extraction to complete to the signature handwriting width information and gradation information withdrawing, has provided the signature handwriting static information support for the further research, finally, used the VC++ platform the MFC frame to complete these algorithms, and manufactured software. At present the software has tried in China Criminal Police University to move, obtained the satisfactory movement effect.