Spatial Vision is a highly accurate, easy-to-use 3D vision calibration software tool with 0.1 sub pixel accuracy. The software works with a pair of webcams or cameras (USB or GigE) to deliver accurate 3D stereoscopic images and videos in real-time or saved for later use.
Spatial Vision
Spatial Vision’s intuitive user interface allows for fast calibration within 30 minutes. The software can be used for 3D measurement and analysis in real-time. The data can be saved as rectified 3D movies for downstream post processing. Spatial Vision provides programmatic interfaces to 3D calibration files for custom C and C++ applications.
Spatial Vision Robotics
A robotics version, called Spatial Vision Robotics, provides real-time object recognition and vision guidance for machines and robots. This provides real-time position and pose information (X, Y, Z, Rx, Ry, Rz) for random box moving, mixed depalletization, random bin picking, random bag picking, and random part inspection. It can work with a wide range of robots.
Key Features of Spatial Vision
The advanced vision processing and accuracy tools of Spatial Vision allow you to save the live camera feed as 3D still images or movies and then use the calibration results for your own application.
About Universal Robotics
Universal Robotics’ signature technology, Neocortex™, is a sensory-motor based form of artificial intelligence that enables moving machines to learn from their experiences and perform tasks that are unsafe or difficult for humans. Neocortex was developed over seven years with NASA and Vanderbilt University, and was funded by U.S. Department of Defense. The company is automating intelligence.
Because Neocortex’s learning requires observation and interaction with the physical world, Universal created an affordable, millimeter-accurate 3D vision positioning product called Spatial Vision™ Robotics that provides Neocortex with information about its surroundings using 3D sensors and cameras.
The company has also developed motion control software, Autonomy, which automatically generates robot motion control code that allows machines to react and move dynamically in real-time to random input.
1 Align Cameras. After you have plugged in your USB 2.0 or GigE cameras, Spatial Vision will find a pair, and give you
simple tests to physically align your pair of cameras. It will also determine left and right channels.
2 Calibrate Space. Next, Spatial Vision will guide you through calibrating your 3D space taking 15 sets of photos using a checkerboard fiducial. This software will indicate which images are adequate for calibration. When complete, Spatial Vision gives you the calibration error, as well as an Estimate Possible Accuracy calculator to determine if your calibration is adequate (if you just want to create a 3D movie, this
won’t be needed).
Total Time = ½ hour!
3 Use Tools. For more precision, you can conduct an epi-polar line test, view the 3D workspace, or go right to viewing and saving your combined 3D anaglyph or Laplacian of Gaussian Edges image or video, either in raw or rectified format. For 3D analysis, you can measure specific X, Y, Z of coordinates or distances.
4 Generate Output. You are ready to create and save either combined 3D videos or images (raw or rectified). You can view your 3D anaglyph image or video, take a 3D snapshot, record a video, replay and analyze your images in 3D. The calibration files available for your C or
C++ applications that describe the warp maps, lens distortions, translation matrices, and calibration error for this configuration of pair of cameras.
There are also color, object and face tracking tools that can be used in demo mode.
Spatial Vision is designed for creating precise 3D calibration, delivering precise 3D position information, and generating visually-accurate 3D bitmaps or movies. Spatial Vision Robotics is an enhanced version that provides real-time precise position and pose in six degrees of freedom of objects for robots.
