With the development of digital twins, the relationship between real space and virtual space has become more and more important. Traditional VR cameras can only obtain flat images, and pure lidar slam can only obtain invisible spatial information. VR The combination with the laser radar slam can obtain a color real-time 3D space model, but the combination of multiple sensors often has many barriers. How can the scanning data of the moving laser radar and the shooting data of the moving VR camera To correlate correspondence? Is there a solution to allow the VR camera to accurately color the slam model? This puts higher demands on the product design framework. The multi-eye VR camera solution based on NT98530 can effectively solve this problem. First look at the technical framework:
1. In this solution, NT98530 can drive 4 sensors synchronously, and the highest resolution can reach 5 million, ensuring the clarity of shooting
2. The 4 sensors can achieve synchronous exposure and synchronous white balance, so that the pictures of the 4 sensors look smooth and uniform, with consistent colors. The difference in brightness and color between the four sensors is avoided.
3. Four sensors can realize multi-dimensional anti-shake of a gyro-sensor, which can avoid hand-held or car roof image shake.
4. This solution can support the hardware-level PTP protocol, which can maintain the frame-level synchronization with the lidar, ensure the pixel-level correlation between the slam data and the camera, and facilitate the real scene coloring of the slam 3D data.
In addition to the above advantages, this solution has low power consumption. The typical power consumption with 4 sensors is only 4 watts. At the same time, there are abundant other peripherals that can be used for other applications.