Editor’s Note: In this technical publish we clarify how Seurat, a developer device from Google, is ready to carry high-end graphics to cell VR methods.

If college students might see virtually anyplace on the earth with the prime quality graphics of Google Earth VR, it will create new alternatives for exploration and studying. For this to work, it will have to be based mostly on a system that’s accessible and straightforward to make use of for faculties. Using a brand new device known as Seurat, we have been capable of just do that. We just lately launched assist for Google Earth scenes in Expeditions, which lets hundreds of thousands of scholars world wide expertise a few of the most lovely locations on the planet.

Mobile VR is straightforward to make use of, but it surely has much less obtainable computing energy than PC-based methods. To make Google Earth VR hit the required framerates on cell VR, we have to use strategies like occlusion culling. Google Earth has one of many largest 3D datasets obtainable at this time, and we’re continuously updating it with new scans and making use of reconstruction algorithms in our knowledge facilities. However, the tempo of updates makes it difficult to use current analysis advances.

Fortunately, Seurat can assist cope with scenes which are closely occluded. We introduced this device at Google IO, and it’s an answer for top of the range graphics on cell VR:


Seurat is a scene simplification know-how that optimizes the geometry and textures in your scene from an outlined viewing quantity (dice to the left above). It takes layered depth photographs (RGBA + Depth) as enter and generates a textured mesh, concentrating on a configurable variety of triangles and texture measurement. Having management of triangle rely is useful for the Google Earth knowledge, the place the standard of the vertex knowledge varies between areas. Hitting constant framerates is necessary for VR, and the extra dependable the scene geometry rendering time is, the simpler it will get to arrange per-frame price range for different components, corresponding to menus and overlays. Below is an instance of producing Seurat enter from Google Earth knowledge.


As seen under, the mesh appears to be like good contained in the viewing quantity. As quickly as we transfer exterior, we are able to rapidly see the consequences of the Seurat occlusion optimization. In this instance, we used hundreds of thousands of triangles and 1000+ MB of texture knowledge to generate the Seurat enter. The output is a 20ok triangles mesh (50:1 compression) with 5 MB texture knowledge (200:1 compression).


In present writing, cell VR solely helps 3DoF (rotation solely) monitoring. 360 diploma photographs, each mono and stereo, have been utilized on this atmosphere with nice success.

Seurat is a robust device for bringing immersive experiences with excessive constancy graphics to cell VR. Hear what our pals at ILM needed to say about it, and if you wish to take a look at these Expeditions with Earth VR imagery for your self, obtain the Google Expeditions app on iOS or Android and luxuriate in them in your Cardboard or Daydream View. Seurat continues to be in improvement, however keep tuned for future updates!

This article sources info from The Keyword