The following is information on MBARI, one of many of their projects and how RaveHD fits into their workflow.
I have to say this is our nerd project and unlike any of the other customers using RaveHD, things are definitely a bit fishy here. Check out their website, they are doing some very cool things: mbari.org
 | Detection and Tracking of Objects in Underwater Video
For oceanographic research, remotely operated underwater vehicles (ROVs) routinely record several hours of video material each day. Manual processing of such large amounts of video has become a major bottleneck for scientific research based on this data. MBARI has developed an automated system that detects and tracks objects that are of potential interest for human video annotators. By pre-selecting salient targets for track initiation using a selective attention algorithm, we reduce the complexity of multi-target tracking, in particular of the assignment problem. Detection of low-contrast translucent targets is difficult due to variable lighting conditions and the presence of ubiquitous noise from high-contrast organic debris ("marine snow") particles. |
 | Onboard one of many of the boats MBARI operates daily
Four times weekly, the research boat Point Lobos takes scientists out to conduct experiments where the ocean is 3,000 meters deep.
Point Lobos is no ordinary boat. Owned by MBARI, it has a laboratory on deck for experiments and a special operations room with banks of monitors that resembles NASA's mission control.
The screens show live pictures from the boats ROV (remotely operated vehicle), called Ventana. It probes the sea with its cameras, data collection sensors, mechanical arms, and animal-collecting devices. A crane on board the boat guides the Ventana into the water. The camera controls are guided from the operations room, using what looks like a joystick for video games. |
 | The launching of a ROV
At the edge of Monterey Canyon, a few hours offshore, goes more than 3,000 meters deep. This means marine scientists can get to deep ocean water and back to shore in one day. On the East Coast, it would take 12 hours to get to water that deep.
The Ventana probe is unleashed. It is guided carefully into the water using the crane fixed on Point Lobos's deck. Floats are layered along a rope to help track Ventana. As it sinks down, its cameras give the researchers in the control room a fish-eye view of the ocean.
The first task takes two hours, filming two-thirds of a mile of the ocean floor at a depth of 200 meters. The distance and depth have to be precise, the tape will be compared to a tape of the same section taken on June 18, 2004. This way, they can find how life on the ocean floor is changing. |
 | The "mission control" center on the ship--the ROV control room. Scientists spend about 10 hours a day in this room, directing the ROV science camera and watching and discussing the geologic features seen. The ROV pilots spend about 14 hours a day in here (in rotating 1-hour watches), since they must be present to fly the vehicle, even during the ROV descent and ascent. The ROV pilot (in this case, Buzz Scott) is in the center; the co-pilot is in the foreground. |
 | AVED is an automated system for detecting marine organisms visible in the videos. Video frames are processed with a neuromorphic selective attention algorithm. The candidate objects of interest are tracked across video frames using linear Kalman filters. If objects can be tracked successfully over several frames, they are labeled as potentially "interesting" and marked in the video frames. The plan is that the system will enhance the productivity of human video annotators and/or cue a subsequent object classification module by marking candidate objects.
This pair of images shows a single frame from an MBARI video taken during a survey of midwater animals. The upper image shows a barely visible siphonophore (middle) and a jelly (lower left). The lower image shows this same frame, in which the computer has picked out these animals from among the marine snow and debris, even though they are barely visible to the untrained eye. Note: These images have been enhanced for publication. The original frame is darker and has even less contrast.
|
 | MBARI has annotated 11,000 hours of underwater video footage from over 300 dives the institute conducts each year. The footage is manually scanned for sea creatures and annotated by experts through VARS (Video Annotation and Reference System) and is available to the public for free through the VARS Query system. An automated system under development called AVED (Automatic Visual Event Detection) uses neural network technology to process the frames . At the moment the current system is capable of highlighting potential objects of interest, automating a very time consuming step. The goal for MBARI is to have an autonomous real-time annotating system. |
 |
RaveHD was chosen to be the ingest point for the video on this project in addition to utilizing the metadata functionality to provide an easy way to locate files. MBARI has been filming since the 70's and the amount of data can be overwhelming. Video is captured directly into a Beowulf cluster that has an attached system that RaveHD sits on. Operators use windows machines using proprietary software developed by MBARI that access and control RaveHD via the net422 protocol provided by SpectSoft.
The project is moving to HD which requires using a DPX file format just like what is used in Hollywood. It is important to store the timecode information and the DPX format allows this to be done in the header, along with additional information that may be needed. The DPX also allows the packing of 8 or 10bit and SD or HD, a great solution all the way around. |  |
|
