Cameras are everywhere these days, observing and recording people, places and things. Most of the time there’s only the one opportunity to capture a moment or an event, and if, for example, the lighting is poor or the camera is out of focus, the quality of the recorded video may render is worthless. This is especially true of video that is to be used as evidence in Court. Video enhancement is a service that can improve the quality of a poorly recorded video.

Forensic vs Non-Forensic Enhancement

Although the work of enhancing an out-of-focus home video is no different to that of enhancing an out-of-focus CCTV recording that will be used as evidence in Court, the way we handle, store and work with each type of video is vastly different.

Video Enhancement Techniques

Technique	What it entails
Color Correction or Grading	Recover the colors of the original scene by compensating for varying filming conditions (e.g. poor lighting). Calibrate video camera using a known standard and adjust color mode, brightness, contrast, luminance, saturation and hue.
Contrast Adjustment or Histogram Equalization	Intensify contrasts. Change the color palette by equalizing the tonal distribution of the image.
De-Interlacing or Motion Compensation	Isolate and analyze separately each interlaced field (with only half the horizontal lines). Combine fields or interpolate missing lines to display full frames.
De-Multiplexing	Isolate and analyze separately individual camera views (scenes) that have been recorded sequentially. Reorder the frames by connecting the ones that appear to be similar in content.
Edge Enhancement or Sharpening	Enhance the apparent sharpness or definition of the video by creating crisp, high-contrast edges. Identify sharp edge boundaries in the frame (e.g. contours between subject and background) and increase the image contrast in the area immediately around the edges.
Frame Averaging	Reduce noise and video graininess. Average each individual pixel from multiple sequential frames.
Homomorphic Filtering	Highlight details obscured by shadows (e.g. licence plate in the dark).Simultaneously normalize the brightness and increase contrast.
Image Segmentation	Locate or isolate elements and boundaries within the frame (e.g. face recognition, plate recognition). Identify edge boundaries and label every pixel within the frame such that pixels with the same label share certain visual characteristics.
Image Stabilization or Tracking	Counteract the visible frame-to-frame jitter caused by subtle camera movements or the motion blur caused by high-speed movement. Distract horizontal and vertical movement or track the target object by slightly shifting the image within each frame.
Image Subtraction or Differencing	Isolate patterns or isolate changes between two frames. Capture lightfield (reference) image and subtract from the input image.
Inverse Filtering	Recover the original frame from a frame that has been enhanced, degraded or corrupted. Reverse the transformation process applied to the degraded frame.
Masking or Blurring	Obscure an area to hide sensitive information (e.g. face, licence plate). Apply a mosaic or blur on the relevant area within the frame.
Noise Reduction	See image or frame averaging. See image or frame averaging.
Photogrammetric or Geometric Correction	Remove shading artifacts and distortions caused by the mapping of non-planar (e.g. 3D) geometric shapes into a two-dimensional frame. Derive the required spatial transformation by analyzing known reference points.
Photogrammetric or Reverse Projection	Derive reliable geometric measurements from a frame (e.g. height, distance, speed). Obtain reference measurements and use a calibrated measurement standard to extrapolate real-world measures.