American researchers used light field, algorithm and other technologies to develop the first robot- specific camera. This single-lens camera can generate 4D images and capture nearly 140 degrees of information. Researchers believe that its performance is superior to the imaging systems currently available for robot vision, as well as virtual reality and augmented reality. The robotics field includes unmanned vehicles, drones, etc. They need to collect information from different perspectives when they are moving around in order to understand the characteristics of the surrounding environment, such as the movement and material composition of different objects.
The 4D camera, developed by researchers at Stanford University and the University of California, San Diego, for robots, can collect a lot of this information in one image.
The reason why 4D cameras can capture more information than 2D images taken by ordinary cameras is due to the use of light field photography. Light field technology was first proposed by scientists at Stanford University in 1996. This technique records the direction and distance of light striking the lens while capturing images like a conventional 2D camera, thereby producing a 4D image.
A feature of the light field camera is that because the image contains information such as the position and orientation of the light, the user can take a picture and then refocus. Robots can use this feature to see the environment through rain and other objects that may blur the line of sight.
Thanks to a specially designed spherical lens, the 4D camera has an ultra-wide field of view that covers a scene that surrounds the camera by nearly a third of its circumference. In the conversion of spherical images into planar images, researchers combined optical technology with advanced signal processing and algorithmic techniques to develop digital solutions that overcome the shortcomings of traditional conversion methods that are cumbersome and prone to errors.
The researchers are nexting to develop a compact robot-specific 4D camera. They believe that small and lightweight, wide field of view, detailed depth of field information, these are ideal for robots, driverless cars, wearable devices and imaging systems for augmented reality and virtual reality.
The 4D camera captures high-quality close-up images, making it ideal for robots traveling in tight spaces, drone landings, and driverless cars. If the augmented reality or virtual reality system is equipped with such a 4D camera, the detailed depth information helps to seamlessly present the real scene and supports a better integration of the real scene and the virtual scene.
Gordon Weitzstein, assistant professor of electrical engineering at Stanford University, who is involved in research and development, says that such cameras can enable different types of artificial intelligence to understand how far, how moving, and what constitutes objects.
1-Piece Fir Tree Cable Tie,Oval Mount Cable Tie,Oval Mount Zip Tie,Oval Head Cable Tie
Wenzhou Langrun Electric Co.,Ltd , https://www.langrunele.com