A combine approach of preprocessing in integrated signature verification (ISV)

Abstract

For last few decades, signature verification is an important area of research. Recently, integrated signature verification (ISV) comes in a play, in which dynamic and static both signatures verified for the forgery [2]. In integrated signature verification system, initial start with a data acquisition stage which can be done from both handwritten and with the use of stylus. Then next step is pre-processing of the signature to make the image noise free and easy to extract. Third and most important step that is the feature extraction. In this step we find that images have different types of features such as local, global, geometrical, and statistical and projection. Last and Final step, which is a crucial step on which the whole system depends that is the verification, where the forgery factor has been found in terms of FAR, FRR, EAR to calculate the performance of the system. Many techniques and filters have been already used to remove the noise in the signature verification system. We proposed a system on integrated pre-processing of the signature. While scanning the signature, some noise is added, which gives the blur image for feature extraction. To improve the system performance and fine feature extraction, we develop a system for integrated pre-processing. In addition, current methods used for features extraction and approaches used for verification in signature systems are also presented. In conclusion, we suggest some encouraging ideas to be incorporated in the future.