Yasmin Nazari*
Identifying on the web anomaly in the video observation is a provoking issue because of real time, video commotion, exceptions and goal. Customary direction based peculiarity discovery model which examinations the video preparing designs for irregularity recognition. This paper means to tackle the issue of video clamour and abnormality recognition .In this paper, a clever separated based group bunching and characterization model was executed utilizing the edge based strategy, diagram based grouping calculation and arrangement model. Trial results demonstrated that the proposed model has high calculation identification rate contrasted with customary constant inconsistency discovery models.