Electronic Design

What's Coming In Machine Vision

Integrated camera vision sensors will continue to evolve, offering even greater image resolution and faster response times that are sure to push the boundaries of data-transfer speeds. Prices will continue to fall, too, making machine-vision cameras affordable for more applications.

At some point, though, there will be a practical limit on much high-performance sensing is needed at the point of inspection. And, processing circuitry located away from the camera sensor and serving multiple cameras at different locations will influence a machine-vision system’s layout. Advances in wired and wireless networking technologies, as well as standards like GenICam, GigE, and Camera Link, will ultimately simplify layout, leading to easier and faster communications.

Look for 3D vision sensing to address the needs of many applications. A number of machine-vision experts point out that viewing an object in 3D space rather than 2D, and then carrying out a complete inspection, will be a major step forward. For example, Vision Components has developed a stereo-type camera with two heads for 3D measurements, robotic guidance, and other applications.

Cognex’s OmniView can unwrap a 3D image to inspect the entire curved surface of cylindrical containers without having to mechanically rotate the container. “This greatly simplifies and reduces the cost of 360° inspection applications in the food and beverage industry,” says Declan O’Dea, manager of Cognex’s U.K. operation.

Although the industrial market is still the largest market sector for machine vision, non-industrial applications are on the rise. These include surveillance, biometrics, traffic and transport, health care, leisure, entertainment, and pharmaceuticals.

The Swiss Institute of Pervasive Computing created software that turns basic cell-phone cameras into an intelligent video-surveillance network using Bluetooth technology. Known as Facet, the software automatically shares information and video between phones and analyzes captured events. Phones can alert each other when any of them detects an object entering or leaving its field of view. Information can also be transmitted to a computer for further analysis.

“Two of the largest future intelligent video applications will be in the consumer and automotive sectors,” says TI’s Flinchbaugh. He foresees the use of cameras in and out of the home, either wired or wireless, to monitor and issue alerts about sleeping and playing infants, the elderly, backyard swimming pools, and front-yard lawns. He also expects a large role for intelligent video systems in lane-departure warning, collision avoidance, driver awareness, and other automotive safety applications.

TAGS: Automotive
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.