As the day to day security is a major concern the authentication problem is very crucial. The hand recognition system provides efficient way to produce the authentication using image processing. Hand recognition geometry as the name suggest uses the shape of the hand to identify the person. Unlike iris, face or fingerprints, the human hand is not unique. The existing systems use finger length, thickness and curvature for the purpose of verification but not for identification. Hand recognition geometry data is relatively easier to collect to other technologies e.g. for fingerprint collection good frictional skin is required by image systems. The main objective of this study is to develop a system which can increase the accuracy of the hand recognition using soft computing, the system should be capable to self-learn, about the correct and incorrect palm prints and add them to its database. A robust palm print recognition approach using neural network is proposed in this study. Neural networks offer a number of advantages, including requiring less formal statistical training, ability to implicitly detect complex nonlinear relationships between dependent and independent variables, ability to detect all possible interactions between predictor variables, and the availability of multiple training algorithms. Palm print recognition, Preprocessing, Feature extraction, Matching and Results and Feedback to the database are the six steps that are followed in the proposed approach.