Survey of Detecting Objects or Facial Images in Real Time Surveillance to Reduce the Crime Rate Using ML & DL Techniques

Detecting and resolving the crime is a very hazardous activity in recent days, but it is possible with the techniques available in machine learning and deep learning neural networks. Surveillance plays the major role in crime detection when a criminal does not belong to any network. The one and only way to find the criminal without network is real time surveillance. Once the criminal's face detected, we need to find and match that with real time surveillance. This study interrogates several articles to spot the various ML and DL algorithms handed down to prediction/detection of objects or facial images which was involved the in-crime scene, and those images were observed in the surveillance. This survey talks about techniques for object detection and Pattern recognition from the crime scene frames and its need to classify the object as a human facial image, animal, vehicle or any other weapon in the crime scene frames with the help of image processing and pattern recognizing algorithms. This is a clear study about what are the data sets and methodologies that were used. Also deals with the powerful algorithms for classifications, Neural Network techniques of ML, DL for detection and the parameters used to detect crimes.