Forensic Video Evidence Authentication

DESCRIPTION:

VDExpert is a powerful forensic video analysis software system dedicated to video authentication and video stream integrity and continuity confirmation. VD Expert is designed for forensic video experts, being a leader in forensic audio technologies and media file authentication.

The main focus of VD Expert is the continuity of the video stream, thus the majority of the methods are dedicated to extraction of specific data from video frame and its following changing during time. Abrupt changes of a technical characteristics can be a trace of video stream continuity break, fraud or editing of video signal.

The approaches to video authentication must include methods of analysis from interrelated fields of knowledge: digital forensics, image fraud detection methods - intra-frame analysis, video stream con-tinuity breaks detection – inter-frame analysis and technical features of the recording device reflect-ed both in video file and video signal.

Technical Specifications:

File structure analysis (correct parsing and format check);

• Information detailing for every header field;
• HEX and ANSI data representation;
• Frame allocation mapping;
• Software traces detection;
• All metadata extraction (XML, GPS, etc.)

Brightness graph analysis:

• Graph of brightness of a selected area (including all frame);
• Dynamic spectrogram of brightness graph (frame size and type);
• Narrowband harmonics detection;
• Energy and phase plots of selected harmonic continuity analysis (frame size, step, initial phase);
• Copy/Paste search in brightness graph (autocorrelation analysis using decision making threshold);
• Display of video copy/paste fragments in two video display windows for detailed comparison.

Date/Time stamp automatic recognition:

• Different date/time stamp shapes recognition algorithms;
• Automatic recognition of date/time stamp values;
• Marking frames with date/time stamp value changing;
• Output of date/time stamp recognition results in table form;
• Plotting of date/time stamp recognition results.

Coding errors of digital image (ELA):

• Calculation and visualization of two-dimensional graph of quantization errors for video frames and images;
• Detection of degree of compression of JPEG algorithm;
• Zoom and navigate in the plane of the quantization errors graph;
• Synchronization of quantization errors graph and the original image in two video display windows

Image colour gradient analysis;

• Calculation and visualization of a two dimensional graph of colour gradients;
• Colour gradient graph calculation: using Scharr operator, using Sobel operator;
• Contrasting of colour gradient using a relay filter;
• Zoom and navigate in the plane of colour gradient;
• Synchronization of colour gradient graph and the original image in two video display windows.
• Display of frame difference in video frames sequence.

File structure analysis:

• Information detailing for every header field;
• HEX and ANSI data representation;
• Frame allocation mapping;
• Software traces detection;
• All metadata extraction (XML, GPS, etc.)

Date/Time stamp overlay image analysis.

DEVICE ID

INTRA-FRAME

PRNU analysis (unevenness of video matrix elements):

• Calculation and visualization of a two dimensional graph of PRNU of the matrix;
• Calculation and visualization of a two dimensional histogram of PRNU of the matrix;
• Contrasting of the graph using histogram;
• Zoom, navigate and synchronization of PRNU of the matrix graph and the original image in two video display windows;
• Automatic detection of range;
• Automatic comparison for different video recorders.