With the introduction of a new research model for Microsoft HoloLens, researchers and developers can now capture a wider range of data captured by the device's sensors.
Application code can now be obtained with the latest Windows 10 update for HoloLens Computer vision algorithms such as Simultaneous Localization and Mapping (SLAM) for spatial mapping and motion, and audio and video streams
"Microsoft HoloLens is the world's first self-contained holographic computer," wrote Marc Pollefeys, a Microsoft research associate, in a blog post. "Remarkably, the research mode available in HoloLens' latest version of Windows 10 is also a powerful computer vision research device."
The data is collected by the device's image sensors, unlike the standard video camera that applications normally access. It works with a quartet of light-sensitive grayscale cameras to track visual features in front of the device and in its periphery. These sensors work in conjunction with the depth camera, which uses infrared light to more accurately measure depth using the time of flight method than standard cameras (which are more sensitive to ambient light).
"Researchers can continue to use the results of the built-in computer vision algorithms, but now can also choose to use the raw sensor data for their own algorithms," wrote Pollefeys. "This opens up a broad range of new computer vision applications for HoloLens."
In addition to processing the data on the HoloLens, users can wirelessly stream the output to a PC or cloud environment for more advanced computing.
For these applications, HoloLens' ability to visualize the results of algorithms in the 3D world before the user can be a decisive advantage, "wrote Pollefeys." HoloLens sensing capabilities can also be very valuable to robotics For example, they allow a robot to navigate through its environment. "
The new research mode has been included with the Kinect for Azure Sensor Project The next generation of HoloLens will be at the IEEE International Conference on Computer on Tuesday Demonstrating Vision and Pattern Recognition (CVPR) in Salt Lake City.