Ikena ISR. Know now.
Current version: 3.4
Inspecting and identifying damage in underwater pipes is time-consuming, and expensive. Pipeline services companies, hired by oil companies, deploy underwater cameras to scan miles of underwater pipe looking for...
Inspecting and identifying damage in underwater pipes is time-consuming, and expensive. Pipeline services companies, hired by oil companies, deploy underwater cameras to scan miles of underwater pipe looking for damage. The company doing the inspection needs to detail the damage and its precise location so that they can direct emergency repair teams to the right location. The telemetry data from the underwater cameras is not designed for pinpoint accuracy, and location information can be off by as much as several meters. This can translate to significantly more search time for the repair teams. Accuracy can be greatly improved by relating the location of the damage to the location of the nearest joint or fitting on the pipe, but the cameras pass so close to the pipe that they don’t capture a wide enough field of view to provide those visual reference points.
Solution: Ikena ISR creates a seamless “mosaic” image where every frame of video is stitched with the previous frames to create an image of the entire pipe, such as the one you see here. With this image product, precise measurements can be derived from the position of the damaged area relative to the nearest joints and fittings.
A far more accurate and clear image of the damage is created and its location is pinpointed to within inches. The emergency repair teams can then be sent to a precise location, and the repairs can be completed more quickly, saving time and thousands of dollars.
Event officials were planning for a their annual two-day festival drawing on average 2,000 people – adults and children. While the festival had never had a serious issue in its 10-year history, the event...
Event officials were planning for a their annual two-day festival drawing on average 2,000 people – adults and children. While the festival had never had a serious issue in its 10-year history, the event organizers wanted to be prepared for any contingency.
The event organizers used Ikena ISR to give them a clear picture of the festival grounds so that they could determine the optimal positioning of their security crew and form an efficient evacuation plan. Ikena ISR enabled them to mosaic the multiple images from their overhead cameras into a full picture of the festival grounds so that the event officials could monitor activity across the entire festival in real time.
Festival officials were able to identify at all times the chokepoints, areas of congestion, and best routes of egress to ensure the success of the festival and the safety of all those involved.
This following describes an experiment done in collaboration with the US Air Force to test an operational scenario
Preparing an attack may take months of preparation. Before and during the attack, the commander needs real-time information on the situation on the ground. The only source of information may be a small drone that supplies quality-challenged video -- in this case, overexposed, interlaced shaky video. In this example, two men are clearly visible on deck, but are the blurry dark shapes people or objects? Missing an armed person could be a life/death mistake.
Ikena ISR’s proprietary algorithms correct the live video in real-time, stabilizing, reducing overexposure, and eliminating interlacing artifacts to reveal that there is indeed a third person on the deck.
The commander is able to launch a successful raid with all threats correctly identified.
Full Motion Video (FMV) suffers from a variety of challenges, improving the quality of FMV requires a suite of sophisticated algorithms to address each challenge. Greater stand-off distances require de-hazing to mitigate the atmospheric effects, and super-resolution to extract more fine details. Small drones are buffeted by winds and lack high bitrate communications, requiring robust stabilization and super-resolution to mitigate noise and compression artifacts.
Most important, to know "what is happening, right now," these capabilities need to operate on live feeds, not on recorded data, after the fact. This is where MotionDSP is different than our competitors: Ikena ISR accepts a live, 30 frame per second video feed (SD or HD), whether analog, digital, or IP, processes it in less than 300 milliseconds, and displays it to the analyst's screen. This is how MotionDSP defines real-time.
Ikena ISR is unique for three reasons: the quality and depth of its image processing algorithms, its speed (real-time), and its use of commercial Graphics Processors (GPUs) from NVIDIA which allow Ikena ISR to operate on any commercial workstation or laptop that meets our hardware requirements.
Detecting motion from a fixed camera (like a mounted surveillance camera) is relatively easy, but detecting motion from a camera that itself is moving is much harder. Ikena ISR’s change detection algorithm can detect moving objects on the ground, from a moving aerial platform, which excellent rejection of false-positives.
Similar to our mosaicing algorithm, it uses Ikena's accurate motion estimation algorithm to model the three-dimensional scene in front of the camera, and then detect and track all the moving objects in the scene. Users can select the level of sensitivity and the granularity of tracking.
Ikena ISR can easily interoperate any Full Motion Video system that support MISB-standard video. Ikena ISR takes an IP video stream or archive file, processes it (real-time fidelity improvement, analytics, and geospatial, or all three at once), and then re-encodes the “fixed” video and displays it to the screen or streams it out as a new IP video stream. Ikena ISR has been proven to interoperate with Jagwire, MAAS, VIDEX, and UVDS. In recent demonstrations, Ikena ISR has been connected to an FMV system in less than an hour.
Small unmanned aerial systems (SUAS) or "drones" are challenged by their small size, their environment, and their communications links. A 5 lb SUAS doesn't have the payload for a stabilized camera ball, and the video suffers from high frequency jitter, their small aperature cameras suffer from noise, and low bitrate data paths result in compression artifacts.
Hollistic Video Fidelity Improvement
Ikena ISR's suite of real-time image processing algorithms address all these issues at once, and in real-time. MotionDSP's massively multi-frame stabilization algorithm corrects the high frequency jitter, adaptive contrast lets you see into the shadows, super-resolution re-constructs the signal removing noise and compression artifacts. All in real-time, in less than 300ms, running on an inexpensive, off-the-shelf laptop running Windows.