in the News

High-def LIDAR Enables Autonomous Cars to 'See Better'

December 18, 2017

One hears a lot about autonomous vehicles these days, but one thing that’s probably not being talked about as often as it should in that context is safety. How can all those driverless vehicles “see” the surrounding environment and avoid crashing into one another?

It’s a concern that TetraVue will address at their booth at CES 2018. The Vista, California-based company has developed what it says is the world’s first light detection and ranging (LIDAR) camera technology that captures real-time images with depth perception down to the pixel level.

LIDAR technology is already the standard for autonomous vehicles, but TetraVue is taking a new approach: merging the multi-megapixel resolution and motion capture accuracy of digital video with LIDAR’s range capability. By illuminating each scene with a non-visible, eye-safe flash at up to 30 frames per second, a TetraVue camera has the ability to process 100 times more real-time spatial and motion data about the surrounding environment. Distance to each pixel in an image is determined through a patented optical encoder coupled with a multi-megapixel CMOS image sensor; the camera then outputs a greyscale, high resolution image with depth information registered to each pixel. AI-enabled advanced driver assist systems (ADAS) can utilize that data to more quickly identify dangerous situations.

The company refers to the new camera technology as adding a “fourth dimension” to digital video, and says that it is poised to transform other markets as well. A better set of high-definition machine vision “eyes,” for instance, could help transform Industry 4.0 retooling of manufacturing and distribution.

In cinematography and AR/VR applications, pixel-level depth information also has the potential to revolutionize motion capture, digital asset acquisition and 4D visualization. By employing multiple cameras or panning a single camera, motion-accurate digital 3D models of actors, objects or scenes can be captured and rapidly turned into virtual worlds — transforming labor-intensive manual approaches and reducing post-production time and cost.

The TetraVue full-motion 4D camera will be on display in the North Hall, Booth 9130 at CES 2018, which runs January 9-12 in Las Vegas.