VanGogh Imaging: The Embedded Vision Alliance Keeps Expanding
In Saturday's writeup, I mentioned that two companies had recently joined the Embedded Vision Alliance, Synopsys and VanGogh Imaging. Actually, as of earlier this week, there are three new member entries; I'm happy to pass along the news that Aptina Imaging is also now a member, and I'll get them added to the member area of the website (along with posting a news writeup like this one) as soon as possible. For now, however, I'm going to focus on VanGogh Imaging.
Greg Werth, the company's Vice President of Sales and Marketing, will be VanGogh Imaging's primary liaison to the Embedded Vision Alliance. According to Werth, VanGogh Imaging provides affordable and easy to use embedded vision solutions for high volume applications that can use mobile devices to capture, display, and measure objects in 3D. Many applications exist, says Werth, in the medical, manufacturing, security, consumer electronics, and entertainment industries that can benefit from applying VanGogh’s embedded vision technology to capture, measure, recognize, and display objects and scenes accurately in real time.
Werth passed along to me flyers on the company's two primary software packages, whose PDF flinks I'll include in the paragraphs to follow. The products' names will undoubtedly prompt smiles from those of you who, like me, are post-Impressionist painting aficionados (or, for that matter, Don McLean music fans).
Starry Night is used to identify objects in a real-world setting and determine their location, orientation, and measurements. The software can analyze static 3D scans as well as real-time 3D streams. Starry Night can also be used to identify human gestures. It works with low-cost, consumer-grade, structured light or simple stereoscopic cameras to render 3D images, and provides very high resolution details based on VanGogh’s patented “surface-based registration” technique. It also works in real time, with GPU-based processing possible, and can be embedded into low-cost laptops, tablets, mobile phones, and Linux micro-PCs. Its very simple interface takes images as inputs and outputs position, size, and orientation in XML format.
Vincent offers users an easy way to embed complex 3D processing functions into applications without the need to be an expert in 3D modeling and computer vision technology. Vincent can run on a number of operating systems including Windows, Linux, Android, and iOS. It add the capability to capture 3D models using a low cost camera solution or on-board camera, such as the Microsoft Kinect, products from Asus, stereo cameras and 3D scanners. Its real-time processing capabilities include depth map generation from stereo cameras, calibration check and calibration correction, cropping, smoothing, and compression, and pre- and post-registration and merging. And 3D models can be displayed on a mobile device using a multi- touch screen.