Here Cloudburst is operating in demanding ambient lighting conditions, inside an automobile. The video shows streaming pointcloud data from Cloudburst. Multiple views are shown from a single capture by simply rotating the pointcloud to a different perspective. This straight-out-of-the camera 3D data illustrates Cloudburst’s low frame-to-frame noise. Sub-mm depth precision enables details, such as user actions and seat belts to be easily identified and they aren’t lost frame-to-frame in sensor noise.
While many depth sensors will only work in a living room, or carefully controlled laboratory, Cloudburst works in rapidly changing dark or bright ambient conditions. As the environment changes, Cloudburst continues to stream 3D data without any additional setup. That’s not just important for automobile interiors, but also for factory floors and logistics warehouses where lighting can be difficult to control.
We’re generating more samples now. Let us know what you’d like to see!