Image vector classification algorithm for hand-writing verification

In this paper, we propose and implement the data mining techniques for verification of hand-writing recorded in an image. The captured images are considered independent of writing material in this system. This system consists of six sub-modules. Namely, i) Sample image data acquisition and preprocessing; ii) Vectors generation; iii) Computation of clusters; iv) Cluster Head Computation v) Pattern Parameter Extraction; vi) Result. The first sub-module captures and categorizes the image for preprocessing. These preprocessed images are vectored and a cluster is computed based on the degree of entropy in the vectors. Therefore, these bunch of clusters represent themselves with the degree of entropy, type of cluster by choosing a cluster head. Finally, the parameters such as the distance, entropy, confidence, are extracted from the clustering; and a result is generated for the given set of samples.