The DLideas series focuses on current developments in the world of DLP® and emerging applications using the technology. In this second video on machine vision, Justin Lemon, DLi’s Technical Sales Consultant, discusses the wide range of computer vision applications and how integrating DLP® Technology into artificial vision systems can optimize results. In part 1 of DLideas 5: machine vision, Justin introduced what 3D machine vision is and the applications associated. We also took a brief look into computer vision and where we can find these systems in our daily lives. Learn more about machine vision in the video below:
Just like 3D machine vision, computer vision has a very wide range of application subsets. In the medical imaging space, computer vision is used for classifying and detecting disease, for dental scanning, 2D and 3D segmentation, 3D human organ reconstruction such as MRIs and ultrasounds, and vision-guided robotic surgery. It is also used for training physicians on medical procedures and as models of the human body to support diagnosis efforts. In the transportation space it is used for autonomous vehicles, smart headlight systems, to read car license plates of cars, as well as driver vigilance monitoring. In the security space it is used for biometrics, such as near infrared retinal and iris scanning, and for fingerprint scanning. Other applications include; video tracking, object pose estimation, image restoration, motion estimation, gaming consoles, and visual computing.
DLP® Technology offers developers the ability to spatially and sequentially control discrete points of light with speed and precision. Unlike contact coordinate measurements of scanning lasers, structured light systems using DLP® can produce non-contact, highly-accurate 3D data in real time, which can achieve measurement accuracy to the micron level. DMDs provide user-defined controls to customize the pattern speed and image sequence specific to the user’s application. They also provide the capability to individually control each micromirror pixel, which allows for dynamic pattern control to optimize results from multiple objects or operating environments on a single platform.