This Small Business Innovation Research (SBIR) Phase I project will demonstrate the feasibility of a high resolution 3D imaging system, based on a new technology that allows simultaneous 3D coordinate measurement and high resolution imagery using the same off-the-shelf CCD or CMOS sensor. Although stereoscopic 3D images and movies have existed since for over 100 years in varying degrees of quality, only recently have technologies been developed that are able capture accurate 3D coordinates of points on surfaces, objects, and structures. Accuracies for single points to less than 1 mm have been achieved, but such systems produce no images and must assemble a collection of single 3D points over time, with significant errors and no motion permitted. This project will demonstrate a high resolution 3D camera that produces real-time 3D imagery and video with sub-mm accuracies at each pixel. The Phase I effort will verify the predicted performance of key components and measure the range resolution using a simplified optical system. Phase II will then construct a 3D camera prototype and demonstrate the promised performance.
The broader impact/commercial potential of this project is that many industries, from aerospace to industrial surveying to movie and game special effects, have a need to record and measure objects and scenes in three dimensions. Current technology used to capture the 3D coordinates of objects and surfaces is slow, difficult to use, and either can only be used on static objects or requires special suits and sound stages with very low resolution. In spite of the difficulty and associated high cost, the value of 3D data is such that its use is growing in many sectors. This project will demonstrate a high resolution 3D camera with the ability to simultaneously acquire 3D data and imagery 100X faster with higher accuracies and simplified operation, which could dramatically reduce the cost of 3D data. While this makes a large impact on the current markets, the capability to have high resolution images of moving objects with 3D coordinate measurements at each pixel enables a large number of new markets such as 3D biometrics, security, cost-effective digital heritage preservation, real-time measurement of 3D trajectories, robotic vision, etc. This is a disruptive technology for 3D data acquisition with tremendous broader potential.
This Small Business Innovation Research (SBIR) Phase I project demonstrated a new method to acquire high resolution 3D images. A prototype camera system was constructed during Phase I, allowing TetraVue to capture 3D images using 6 megapixel (Mpx) image sensor. This means that 6 million distance measurements were acquired simultaneously, with each distance measurement accurate to several millimeters. The result is a 6 Mpx image of an object that also includes all shape and distance information. This was accomplished for distances up to 4 m, with similar performance possible to 20 m. Design elements for further performance improvements were also identified during this project along with potential software algorithms that could be used to reduce the cost of a future product. This advance is significant because although stereoscopic 3D images and movies have existed for over 100 years, only recently have 3D devices been able to capture the actual measurements and shapes of objects. For example, 3D laser scanners and range finders can measure distances to distant objects to within 1 – 5 mm accuracies one point at a time. Other 3D devices use triangulation or other optical techniques to determine object shapes to < 1 mm, but only for close objects. Current 3D sensor arrays are limited to approximately 100 x 100 px. Compared to 2D digital photography, all of these techniques are slow, costly, and limited in capability. The approach demonstrated in this project captures 3D images and measurement coordinates quickly in much the same way that 2D cameras are currently used. Industries, from aerospace to industrial surveying to movie and game special effects, will benefit from the capability to quickly record and measure objects, motion and scenes in three dimensions with high resolution. Current technology, e.g. 3D laser scanners and motion capture systems, can only be used on static objects or requires special suits and sound stages with very low resolution. In spite of the difficulty and associated high cost, the value of 3D data is such that its use in 3D industrial survey has been growing at 30 – 40% per year, reaching $425M in 2008. The high resolution 3D camera based on TetraVue’s "light slicer" technology has demonstrated the potential to increase the acquisition speed by 100X over current solutions while increasing resolution and reducing total data collection and processing costs by 10X. While this makes a large impact on the current markets, the capability to have high resolution images of moving objects with 3D coordinate measurements at each pixel enables a large number of new markets such as 3D biometrics, security, forensics, cost-effective digital heritage preservation, real-time measurement of 3D trajectories, and robotic vision.
