Investigating and Auditing Through Facial Recognition Using LBPH Algorithm

Abstract

Technology is constantly advancing, and many individuals submit various video material to social networking websites such as YouTube or Facebook. Since it has become a source of income for many individuals across the world, it is becoming increasingly vital to utilize it in situations when you need to discover a specific person in several video recordings. Another option is to manually go through each video and try to discover the individual segment in order to extract it. Manual searching can take a long time, and it's practically difficult to find a specific video in which a person appears. A person can only focus for 20 to 30 minutes on average to recognize or identify the person in the video, and a video stream may take much longer. Due to the huge quantity of data gathered in the multimedia application, such as videos, a human conducting a video search manually may be difficult to do so properly in such cases. It is critical to automate the procedure in order to eliminate human error and the time it takes to identify the individual in the video footage. Given its popularity and use in applications ranging from our mobile phone games to high-end computers for future forecasting, artificial intelligence may be used to address many difficulties, including this one. In this work, a pre-trained facial recognition classifier known as the Linear Binary Pattern Histogram (LBPH) is utilized to recognize a person in video clips and provide recorded proof of each video in which a person appears, along with the time stamp of his or her visibility. Here, a method is proposed for identifying and tracking down a missing individual utilizing massive video data and Artificial Intelligence without the need for human participation.

Keywords

• Investigation
• face detection
• Distributed processing
• Local Binary Pattern Histogram (LBPH) Face Recognizer Algorithm