Recently in most banks (especially in Africa), there exist various cases of banking staff denying services to customer due to signature differences. This has resulted to a lot of misunderstandings, insults, quarrels, and even losses to most financial institutions. This work presents digital signature verification system to enhance customer services in the banking industry, with the aim of improving the staff customer relationship within the banking domain. This will be developed using image acquisition tool, image processing tools and machine learning (clustering technique). The system will be implemented using mathlab as the software development too. The accuracy of 98% was recorded as the system was validated using a prepared testing and training image containing various real and forged signatures.