Part of the Tri-Suite14 software package, Video Investigator® is designed to process video files and still images alike. There are a variety of methods that can be used in either scenario, which is what makes Video Investigator® such a powerful tool. No other software will offer the vast array of filters and features that Video Investigator offers its users.
1. Video Investigator’s Active Interface GUI enables the user to execute the integrated plug-ins in a highly interactive mode with real-time results (*) which are seen in a video preview. This advanced video processing interface allows the user to find the most optimal set of video processing parameters in a highly interactive way.
2. Video Investigator’s Active Interface (*) provides the user the capability to adaptively estimate the best set of processing parameters for any part of a video frame and vary these parameters if needed from frame to frame, or from a set of frames in a time-interval to another time-interval, yet processing the entire video in a single processing pipeline. This frame-adaptive video processing (*) feature accelerates the user’s ability for time-space-adaptive processing of the entire video without first extracting each individual frame and individual parts of frames. An example of using the frame-adaptive video processing is when debluring the video in which different objects and subjects are blurred differently due to the time-varying motion blur and space-varying defocus blur.
3. The Movie Controller provides advanced video playback with audio support allowing the enduser to adjust which frames of video to playback, how frames are played back, and even synchronize multiple videos.
4. Image sequence compilation converts a set of images into video for analysis or court room presentation.
5. The Record function (*), when enabled, records every action the end-user makes and produces a script that can replay the exact actions and settings the end-user used. Unlike traditional report functions that only produce a PDF of parameter values, Video Investigator® Record scripts can be replayed at any time allowing the end-user to see exactly how the result was created.
6. The Zoom filter resizes images with confidence with several innovative zoom algorithms that retain the highest clarity and edge definition (#).
7. Several images or video can be combined together in layers to form a video composition (*) where every element can be relatively repositioned in the window and in time.
8. The Command Window provides an alternate entry point for executing commands for those who prefer to apply filters in the command line.
9. The Highlight filter focuses the viewer on a user defined area of interest.
10. If required by law, portions of images can be tracked and concealed or made unrecognizable to cover sensitive information with pixilation, blur or histogram adjustment.
11. The RGB Profile Plot plots the line scans of RGB channels separately, allowing for independent 2D numerical visualization.
12. The 3D Profile Plot displays a 3D plot of the entire image or video in time with a plot elevation corresponding to the image intensity allowing for 3D numerical visualization.
13. The Elevation Plot allows for the numerical visualization of images as 2D plots of linear image slices.
14. The Matrix Tool provides a detailed quantitative look at the pixel values in an image.
15. The Video Scene Integrator provides a method of manually sorting video frames to make event timeline (e.g. following a subject from camera to camera).
16. The Component Information Window provides a detailed description of all images and video in Video Investigator including image dimension, color space, data format, and other important metadata.
17. A large assortment of denoise filters (#) provide both traditional and the most advanced methods for removing image noise without reducing clarity and features in evidence video (e.g. sensor noise and poor weather conditions).
18. Deblocking (#) reduces compression blocking and ringing artifacts related to harsh compression without reducing clarity and preserving fine features.
19. The pattern removal filter (#) is an advanced filter that removes pattern noise from images such as latent prints on textured surfaces.
20. The Average Frames filter generates an image by combining together multiple frames of the movie sequence, removing dynamic noise from stationary objects in the video.
21. The FrameFusion® Denoise filter (#) is a state of the art method of removing noise from an image or video with Cognitech’s patented algorithm that preserves sharp edges and fine textures imbedded in noise.
22. The Median filter removes pixels that deviate from the median pixel value such as salt and pepper noise.
23. The Remove Extrema filter is an advanced filer that removes spurious shapes from an image that are due to noise distortions.
24. The Total Variation Denoise filter (#) is a general method, discovered by Cognitech, for removing noise from an image using the Total Variation Minimization technique; retaining edges and minimizing visual artifacts associated with any image processing.
25. A large assortment of deblur filters options provide tools for every type of blur problem including motion, defocus, optical, Gaussian averaging, and custom.
26. The Adaptive Blur filter (#) computationally estimates the blur from a blurred image and allows the end-user to remove complex non-linear blur from any evidence image leading to highly accurate blur removal.
27. The Motion Deblur filter (#) automatically determines the motion direction and length from velocity estimation and removes the estimated motion blur.
28. The Sharpen filter is a simple enhancement solution for producing sharper, clearer images. 73. The Wiener Deblur filter applies any of the available deblur kernels using the Wiener method for fast and accurate blur removal.
29. The Total Variation Deblur filter (#) is a general method, discovered by Cognitech. This filter applies any of the available deblur kernels using the Total Variation method in order to retain edges and minimize visual artifacts, such as ringing, which are commonly associated with traditional blur removal.
30. Automatic histogram equalization provides three different methods of automatically enhancing images and video to obtain the best histogram levels locally or globally (e.g. Morpho-Local Equalization (#) will equally enhance features in both under and over exposed regions of the image and in the presence of haze degradation).
31. The Manual Histogram Editor provides user-defined adjustments for better visualization of low dynamic range content such as night time video.
32. Color correction can enhance hue and saturation of video.
33. The FrameFusion® Reconstruction filter (#) is a Cognitech patented Super-Resolution technique that not only removes noise from video but actually integrates adjacent frame information to form a higher resolution result and works with moving subjects unlike the traditional Video Averaging filter (e.g. license plate number clarification).
34. The Stabilization Reconstruction filter utilizes Cognitech’s advanced velocity calculation algorithms to accurately track and stabilize an entire video or particular object (e.g. vehicle dash cam and handheld camera footage).
35. The FaceFusion3D Reconstruction filter (#) is the newest Cognitech patented technique that combines the principles of FrameFusion® with 3D facial models allowing for the SuperResolution reconstruction of images and videos of human faces in full 3D rotational motion.
36. The Mosaic Reconstruction filter combines multiple frames into a single panoramic picture by accurately stitching common information between frames.
37. The Reconstruction Wizard (*) automates the video stabilization and mosaic reconstruction process for standard use cases.
38. Interlacing and de-interlacing functionality allows for lossless conversion between interlaced and progressive video.
39. The Frame Align operation aligns interlaced frames that are recorded at sequential points in time, doubling the resolution of the objects in the image.
40. The Geometric Lens Correction filter (#) corrects for geometric lens distortions automatically without any knowledge of the source camera.
41. The Fourier filter computes the 2D discrete Fourier transformation of images for frequency representation.
42. The Geometric filter provides an interactive interface for any perspective transformation of video frames.
43. The Mirror Image filter is an interactive tool to rotate an image or video 90, 180,-90 degrees and flip horizontally or vertically.
44. The Automatic Rectification filter rectifies an image to correct a perspective distortion due to camera angle, based on user-defined points.
45. The Manual Rectification filter provides an interactive 3D graphic interface for real-time perspective correction.
46. The Wavelet filter computes the 2D dyadic wavelet transform for spline wavelets of minimal support on a rectangular domain for advanced mathematical analysis.
47. The Correlation Matching filter estimates the global motion between any set of paired frames (motion patterns) of a movie and produces accurate velocity vectors used to properly account for motion, which is utilized in FrameFusion Reconstruction, Stabilization Reconstruction, etc.
48. The Shape Matching filter (#) estimates the dominant motion between consecutive frames of a movie and produces velocity vectors used to properly account for motion, which is utilized in FrameFusion Reconstruction, Stabilization Reconstruction, etc. Shape matching will work when multiple object motions are present in the video (e.g. foreground and background motion).
49. The Interactive Manual Matching filter (*) manually matches motion between any set of paired frames (motion patterns) of a movie and produces velocity vectors used to properly account for motion, which is utilized in FrameFusion Reconstruction, Stabilization Reconstruction, etc. Manual Matching will always work when automatic methods may not work due to the complexity of motion.
50. The Segmentation filter (#) provides an image analysis technique which automatically partitions an image into regions of similar intensity/color used for extracting tattoos.
51. The Edge Detection filter allows the end-user to extract edge features for extracting vehicle markings or license plates in a video.
Supported File Formats and Accurate Video Playback
52. Most image and video file formats are supported including: COG, CFF, 3GP, 3G2, ASF, AVI, BMP, CAL, CMP, CUR, DCM, FLV, FPX, GIF, ICO, IMG, J2K, JPG, M2TS, MAC, MKV, MOV, MP4, MPEG, MSP, MTS, MTV, OGV, PCD, PCT, PCX, PGM, PNG, PSD, RAS, RM, SCT, SWF, TGA, TIFF, TS, VOB, WBMP, WEBM, WMF, WMV, WPG.
53. Streaming video is decoded and converted into a frame accurate format that can be processed and analyzed frame by frame.
54. Cognitech’s patented lossless COG (#) encoding provides at least 50% file size reduction to raw video without any loss of content or quality and is 100% reversible.
55. Cognitech® CogPlayer provides proprietary lossless COG (#) and CFF playback on any computer with Windows Media Player, which enables court presentations on any computer.
56. 32-bit floating point (*) image and video content is supported for more numerical accuracy. All filter computations are done in floating point, maintaining the highest level of quality throughout the signal chain.
57. Cognitech® Video Investigator® and VideoActive® provide frame based video playback allowing for easier and more accurate frame-by-frame analysis than traditional time based video playback software.
*Above items marked with a hashtag (#) indicate unique forensic video capabilities that are proprietary ONLY to Cognitech, Inc. due to issued and pending US and International patents of Cognitech, Inc.
*Above items marked with an asterisk (*) are designated computational design capabilities that are the leading edge of forensic video technology and are not commonly available to other competitors in the field.