Passive Video Tampering Detection Using Noise Features

Introduction

In recent years due to presence of sophisticated video editing software, network technologies, low cost multimedia devices and wide adaptation of digital multimedia coding standards, video tampering has become very easy. Due to digital nature of video file, it can be easily manipulated, synthesized and tampered numerous ways. Additionally there are no requirements of better technical skills to tamper the video. Internet is providing different tools and software without any cost, which is further increasing the ease of video tampering. People do video tampering to hide or expose some important scene or event which was actually not present in the original video. Someone can create a fake video of any popular person and can defame them. Sometimes, Hollywood and Television celebrities are victimized by video tampering attacks. People are also using tampered video to get justice in their favor from court of law. Video tampering has proven the concept “Seeing is not believing”. So it is important to bring out the truth to the world.

Video tampering can be categorized in spatial tampering and Temporal tampering domains. In Spatial domain only a single frame is affected with tampering, but in Temporal domain many frames are affected with tampering. Video tampering detection techniques are basically categorized in two categories, first one is Active video tampering detection technique and second one is Passive video tampering detection technique. In starting, people were using Digital Watermark (DW) and Digital Signature (DS) for video tampering detection which belongs to Active video tampering detection technique. Active video tampering detection technique requires pre embedding information in the video, but if the video does not contain pre-embedding information like DS and DW then it is very difficult to detect tampering in such video. Currently, the main challenge is to detect tampering through Passive video tampering detection technique which does not include pre-embedding information like Active video tampering detection technique. Passive video tampering detection technique utilizes the intrinsic properties of video like noise feature, camera response function, color filter arrays etc. In Blind and Passive video tampering detection technique we have no information of source from which the video has been taken and video does not contain any information like DS or DW, so Blind and Passive video tampering detection technique is much more complex in comparison to active video tampering detection technique.