The main problem with bsv is that we need the people in person to verify in Banks. You know people who are just above 50 years of age is not having the clear lines in their thumbs. So when they swipe their finger it does not operate properly. Even in educational institutions ( Private Colleges) it is mandatory for the employees to swipe their thumb in the m/c. But, many a time i find people were keep on pressing since they do not hear the beep sound. This is true for elderly people. More over we do internet banking only now a days. Banks like SBI does not allow you to withdraw money through the teller counter. Instead they ask you to uase ATM only. It is partially technical problem and partially our attitude and lazyness.

Thank you ma'am ! But i was mainly interested in knowing open issues of biometric signature verification which doesn't suffer from that swipe drawback mentioned by you. 

You might be interested in a recent publication by Fiebig et.al., describing implications of high-resolution smartphone cameras on the security of an individuals fingerprints or iris.

https://www.usenix.org/system/files/conference/woot14/woot14-fiebig.pdf

Another important aspect is that you can't revoke a biometric "token". In case your central authority is attacked, biometric identification values are leaked and can be used to replicate biometric features. And you don't like to cut of the fingers of all customers to solve this issue. There are already reported incidents of fingers cut of but bank robberies to fool the biometric identification.

It is also to consider that most biometric scanners can be fooled with contraptions. I know of at least two studies that showed that all finger print readers can be fooled with hot glue copies. And we don't need to talk about the liveliness analysis ...

The issues as described by Chandramathi Soundararajan are real life social and practical issues. The do invite our attention for finding the solutions technically. They may possibly be considered open issues in online biometric signature verification.

I advise you to read some of the publications of Prof. Hanmandlu of IITD in this regard.

They may be of some help to you.

ok sir, thank you ! Actually signature verification is behavioural biometric but not physiological biometric so issues are different.

I think a challenge may be to be able to support handwritten signature verification on common mobile devices. Furthermore, from a security and privacy viewpoint, it would be very important to avoid to create a big database including users' biometric data for security reasons.

Hope that my last publication on such a topic (attached) may be found interesting for you.

Article A new system for secure handwritten signing of documents

Online system uses stylus to extract dynamic information such as biometric, feature recognition and signature verification etc.,

http://ijarcsms.com/docs/paper/volume3/issue1/V3I1-0056.pdf

I think the best way to identify Signatures and Handwritten characters is to bound them by close contour (using an efficient edge detector; e.g. LOG edge detector), then following the boundary points (utilizing for example the chain coding method) and measuring the moments and used to recognize and identify the desired signature. Another effective method is by using horizontal and vertical projections to define the handwritten or signature boundary. The complex or real moments can be used to identify them