Here is your first look at the editorial content for the upcoming Spring issue of Vision Spectra.
Broad View of the Sensor Market for Machine Vision, Including SWIR
This article presents a broad view of the sensor market, highlighting emerging technologies and applications that are driving the market. Sample text: Image sensors, part of every smartphone, may seem like a ubiquitous, mature technology – however, this is far from the case. While conventional CMOS detectors for visible light are well established, there are extensive opportunities for more complex and innovative image sensing hardware that offer capabilities beyond simply acquiring the intensity values at each pixel. This article, from analyst Matt Dyson, looks at the sensor technology at the center of numerous trends, from autonomous vehicles to unmanned aerial vehicles, to thin, flexible biometric imaging, precision agriculture, and the use of hyperspectral and multispectral imaging in smartphones.
Key Technologies: SWIR sensors, CMOS sensors, hyperspectral imaging. Key applications: Industrial sorting, smartphone imaging, precision agriculture, direct x-ray imaging, medical x-ray, silicon photodiodes, biometric imaging, event based vision, thin-film photodetectors. QD-on-CMOS hybrid detectors.
Automotive Symbol Inspection
Automobile industry standards require that illuminated vehicle elements such as warning lights meet strict specifications for luminance, chromaticity, clarity, and shape. These elements must be tested for safety and performance which is an increasingly complex prospect. Typically, machine vision cameras are used to inspect backlit symbols, matching characters, shapes, and symbols compared to a golden sample. However, they cannot always measure symbols that are misaligned, nor can they measure absolute luminance and chromaticity, which is where photometry-based systems come in. This article discusses visual inspection of backlit symbols, and how the strengths of machine vision and photometry combine for an optimal solution.
Computational Imaging
High-quality cameras increasingly depend on extended systems of intricately optimized aberration-correcting lenses. Unlike conventional lenses, meta-optics leverage subwavelength scatterers to replicate complex aspheres in an ultrathin, flat surface that is mass manufacturable. A fundamental limitation of meta-optics, however, is their strong chromatic aberration. While some meta-optics are achromatic, these designs are fundamentally limited to a small aperture, low NA, and a narrow field of view. Though meta-optics are ill-suited to directly focusing broadband light, their subwavelength spatial resolution and inherent freeform capability enable them to transform light into different bases. These stable bases preserve image information and are used to computationally reconstruct a high-fidelity image. By optimally matching the meta-optics with the image processing routine, reductions in size and complexity are simultaneously realizable in cameras producing high-quality images. Tunoptix Inc., a Seattle startup, is commercializing this technology to facilitate high-performance, simplified cameras for an array of consumer, industrial, space, and medical applications.
Key Technologies: meta optics, image processing, computational imaging
Line Scan Cameras
Line scan cameras image the world a pixel-width at a time, capturing items at high resolution as they move past – often at high speed. Favored for inspections, this approach is like other vision technologies – the choices made have to match the application. This can mean going with near contact imaging for a large, flat surface to something that stands further off when conditions aren’t so planar. And increasingly, it may mean going to the non-visible, with SWIR line scan cameras, for instance, projected to double their market share toward the end of this decade. Examples show when and which line scan cameras should be used.
Key Technologies: line scan cameras, SWIR line scan cameras