Machine Vision: Finding Trends That Matter
manufacturingtechnologyinsights

Machine Vision: Finding Trends That Matter

By Manufacturing Technology Insights | Monday, November 18, 2019

Machine VisionThe expansion of components for 3D imaging in the machine vision market is a definite trend that is being driven by high demand for 3D measurement and guidance, and by the increased availability of cost-effective technologies that are part of 3D imaging systems.

FREMONT, CA: Most trends in machine vision derive from proficient and practical new or evolving technologies. However, in engineering discipline like machine vision, a further consideration that seems suitable is whether a trend is based on existing practical use cases or on extreme forward-looking predictions of future potential that are not yet fully realized. Below are recent trends that might impact the selection and execution of the technologies for machine vision in industrial automation.

• 3D Imaging for Metrology and Vision-Guided Robotics (VGR)

The expansion of components for 3D imaging in the machine vision market is a definite trend that is being driven by high demand for 3D measurement and guidance, and by the increased availability of cost-effective technologies that are part of 3D imaging systems. Part of the development is a proliferation of algorithmic competence for specific applications such as 3D measurement (metrology), robotic guidance, and Autonomous Mobile Robots (AMR) guidance and safety.

Three-dimensional imaging systems capture a sight of physical space and offer data that represent points in the scene comprising depth as well as familiar 2D planar (x and y) locations. A few available components also proffer a grayscale (contrast) or a color image along with the 3D data. The fundamental advantage of 3D imaging lies in providing a 3D location, but a further supplementary benefit is that the 3D image is usually “contrast intolerant.” That is, the picture information lets the software to process depth changes rather than an alteration in surface color features or shadows.Machine Vision

• Embedded Imaging

One definition that classifies embedded vision as any tool that combines image capture and processing, cover a broad and perhaps overlapping segment of what also are conventional machine vision components. One might alternately constrain embedded vision to those devices that are fully integrated at a lower level (Systems on a Chip (SOC), System on Modules (SOMs), or single-board computers) to be integrated into a larger device. Easy to imagine use case examples may include self-driving automobiles and AMRs or even smartphones. In machine vision, the term can also apply to an emerging development where cameras contain embedded vision processing on-board to carry out the application-specific tasks in the camera, rather than on an external computer.

• Advanced Lighting Techniques and Processing

Trending in illumination components for machine vision is the accessibility of controllable, multi-spectral devices that enable more elasticity and superior competence in specific imaging situations. By changing monochromatic colors, one can overcome part family variations without several lighting devices, or even create a color image using various images of different illumination colors. Additionally, high-rate imaging of multiple views using diverse lighting angles could be used to create 3D representations of an object or to offer a High Dynamic Range (HDR) image.

• Liquid Lenses and High-Resolution Optics

Two clear technology directions in machine vision optics are the proliferation of more advanced large-format and high-resolution camera lenses and the move toward further seamless integration of component liquid lenses. The first development is practical; as camera resolutions increase and pixel sizes decrease, the requirement for better optics is understood and being met by most component lens providers. More product specifications comprise details of lens performance, for instance, charts showing the system’s Modulation Transfer Function (MTF), a superior measure for lens comparisons.

Liquid lenses are equipment that can change focus depending on an external without requiring any mechanical change in the lens as in manually focusable lenses. Liquid lenses have been used in smart cameras, smart sensors, and other machine vision devices for years. Conversely, recent advances in the incorporation of the tools with machine vision optics and cameras are bringing the technology more into the sphere of general-purpose use.

Weekly Brief