Identifying Fradulent Behaviour And Its Types using Clustering Techniques

Classification and Clustering, plays a major role in tracking of multiple humans, this is a demanding problem in composite situations in which extended occlusion, shadow and/or reflection exists. This method includes the stages, analyzing, detecting, validating and tracking. The proposed work is concentrated using Density based k-means clustering algorithm, where after detecting and tracking, the level of crime zone is analyzed, since criminal is clustered over one zone to another they are tracked using filtering method. Density based algorithm works strongly in very demanding Sequences. Information stored in database is analyzed using data mining techniques pattern discovery, evolution analysis.