The project was to enhance the security in the election booth. Human detection is done and the alarm goes off when the camera detects more than one person standing in the booth. The model is trained after extracting features from various datasets. Features that signify human figures were extracted and the model was trained. The features were then optimized from the previous results obtained.