Two years ago, I wrote about my growing interest in "embedded vision", the incorporation of computer vision capabilities into embedded systems, enabling those systems to extract meaning from image and video inputs. Though it's not usually spoken of as "digital signal processing," embedded vision typically uses algorithms familiar to digital signal processing engineers. These include algorithms to improve the quality of capture images (such as lens distortion correction, often performed using bilinear interpolation), algorithms to extract features from images (such as the Canny edge detector, which incorporates a Gaussian filter), and algorithms to track objects (which often incorporate Kalman filters).
This strong link between embedded vision and digital signal processing is a good thing, in several respects. Perhaps most importantly, now that embedded vision is being implemented in many cost-, power-, and size-constrained systems, engineers are harnessing well-established techniques for efficient implementation of digital signal processing algorithms in order to squeeze demanding vision algorithms into modest processing resources. In addition, the link between DSP and embedded vision means that companies developing embedded vision technology and products can leverage the expertise of experienced DSP engineers to get their products to market quickly.
The potential for embedded vision is huge. Huge in terms of the value that can be delivered to users of many kinds of electronic equipment, by making that equipment more intelligent and responsive. Huge in terms of the business opportunities for providers of systems, sensors, processors, algorithms, software and tools. And huge in terms of the opportunities for engineers to do exciting work that delivers compelling value to users.
And it is truly exciting to see what can be done with embedded vision, from saving lives (the 2013 Mercedes Benz A class car, for example, is bristling with multiple vision-base safety features) to making systems easier to use via gesture user interfaces. The more you see what people are doing with this technology, the more you'll find yourself coming up with ideas about what you can do with embedded vision in your designs. And the more you understand how embedded vision technology works, the more equipped you'll be to be able to incorporate these capabilities into your products.
To help engineers envision the possibilities for embedded vision in their designs, and to help them gain the practical skills required to make those ideas a reality, last year my colleagues and I at BDTI founded the Embedded Vision Alliance. The Alliance's first project has been to develop a web site, www.Embedded-Vision.com, which delivers a rich stream of inspiring product and technology examples, as well as practical technical training and other resources for engineers who want to understand how to incorporate embedded vision into their products.
And now, I'm very excited to announce that the Embedded Vision Alliance will be hosting its first conference for engineers, on September 19 in Boston. The Embedded Vision Summit will provide a technical educational forum for engineers, including how-to presentations, demonstrations, a visionary keynote presentation, and opportunities to interact with Alliance member companies.
If you're an engineer involved in, or interested in learning about, how to incorporate embedded vision into your designs, I invite you to join us at the Embedded Vision Summit on September 19. Space is limited, so please register now. To begin the registration process for the Embedded Vision Summit, please send an email to summit@Embedded-Vision.com containing your name, job title, and company or organization name. We will respond with further details via email.
Jeff Bier is president of BDTI and founder of the Embedded Vision Alliance. Post a comment here or send him your feedback at http://www.BDTI.com/Contact.
Add new comment