Vision Systems Design Announces 2021 Innovators Awards

July 28, 2021
For the seventh straight year, Vision Systems Design held its Innovators Awards program, which celebrates the disparate and innovative technologies, products, and systems found in machine vision and imaging.

Dennis Scimeca and Chris Mc Loone

Based on the impartial ratings assigned by our panel of judges, awards were assigned in four different levels—Bronze, Silver, Gold, and Platinum, to companies or organizations that demonstrated excellence in a product or technology, an application, or in research and development. Categories to distinguish our honorees include vision systems, four different camera categories, sensors, software, embedded vision, robotics, and more. The 2021 Innovators Honorees are listed below.

Bronze Honorees

ADLINK Technology

NEON-2000-JNX: NVIDIA® Jetson™ Xavier NX-based Industrial AI Smart Camera for the Edge

ADLINK Technology (Taipei, Taiwan; www.adlinktech.comNEON-2000-JNX industrial smart cameras feature the Jetson Xavier NX, a GPU-based System-on-Module designed for embedded and edge AI applications. The cameras support six types of image sensor and feature FPGA-based digital input/output and an optimized OS with vision software suites, all in a compact chassis designed to make edge AI easy to implement.



Shelf-OS features cloud-first open architecture and runs patented algorithms and proprietary software on edge devices called "Shelf-Hawks" that sit atop or underneath supermarket shelves. Shelf Hawks use Intel RealSense cameras to capture on-shelf product status by measuring 3D volumes. Retailers use Shelf-OS for automated workflows to restock shelves faster, while distributors can use the information to track stock levels and optimize delivery routes.

Edmund Optics

TECHSPEC® Cw Series Fixed Focal Length Lenses

TECHSPEC Cw Series Fixed Focal Length Lenses are waterproof lenses designed to eliminate the need for a protective cover over machine vision systems in harsh environments like high pressure, high temperature water spray downs, or automotive and security applications. The lenses integrate directly with waterproof cameras and include a coated window to prevent water droplets from settling on the surface of the lens.

Instrumental, Inc.

Discover AI

The Instrumental platform is designed to capture images of a unit at every production milestone and upload those images to a secure cloud-based solution. Discover AI analyzes the entire build process to discover any deviations and create heatmaps of potential problem areas. By categorizing each potential issue separately with real-time reports, engineers can use Discover AI to pinpoint the root causes of product issues.

LUCID Vision Labs, Inc.

Helios2 Time-of-Flight Camera

The Helios2 Time-of-Flight Camera produces 3D depth data with sub-millimeter precision and processes the data on-camera. It features an IP67-rated enclosure with lens protection for deployment in tough industrial environments. The GigE Vision interface includes Power-over-Ethernet and the industrial M12 connector supports cable lengths up to 100 m. The Helios2 also features improved optics for higher light collection and better calibration for higher accuracy.

Neousys Technology Inc.


The NRU-120S computer is powered by the Jetson AGX Xavier System-on-Module and delivers up to 32 tera operations per second (TOPS) while drawing only 30 W of power. It features fanless cooling and a compact design to support AI inference applications at the edge. It can operate between -25° and 70° C and features four GigE ports and a dual front-accessible hard disk drive/solid state drive.

OmniVision Technologies, Inc.

OAX8000 AI-Enabled Automotive Application-Specific Integrated Circuit

The OAX8000 AI-enabled automotive application-specific integrated circuit is optimized for stand-alone driver monitoring systems such as eye gaze and eye tracking. It integrates 1GB of DDR3 SDRAM, a neural processing unit, and an image signal processor to provide speeds up to 1.1 trillion operations per second. This ASIC and any range of companion OmniVision automotive image sensors draw less than 1 W of power combined.

Radiant Vision Systems

VIP (Vision Inspection Pack) Software

VIP, or Vision Inspection Pack Software offers a complete solution for the inspection of backlit symbols such as those found on dials in a car or buttons on electronics. The software uses photometric measurement of light and color in conjunction with machine vision-based registration of 2D characters and shapes to provide an efficient single-camera/software solution for lab testing and production-level quality control.

SensoPart Industriesensorik GmbH

Target Mark 3D Detection Principle

Target Mark 3D detection technology is designed for machine loading and pick-and-place applications using mobile robots. A coded mark is attached to a machine that a mobile robot is tending. If the mobile robot should be repositioned relative to that machine, the robot can use the coded target mark as a point of reference to calibrate its approach and regain the appropriate movement path for tending the machine.


SICK InspectorP61x - Ultracompact Embedded 2D Vision Sensor Powered by Deep Learning

The SICK InspectorP61x is an industrial ultracompact 2D vision sensor with 1.2 megapixel image resolution and built-in image capture tool. It is designed to support the deep learning process. Integrated optics and illumination can tune brightness, color, and shape of light to obtain high quality images. The included SICK Nova SensorApp allows expert and non-expert users to configure the sensor through an intuitive web interface.

Smart Vision Lights


The LMX75-UVC compact linear light combines 8 high intensity LEDs for machine vision applications, three 265 nanometer UV-C LEDs with transmissive silicone optics for disinfecting surfaces, and an internal driver with OverDrive strobe operation to simplify component count and ease application retrofits. Optional aluminum or nickel-plated enclosures provide IP65 washdown protection to support use in food inspection applications.

Teledyne e2v


The Hydra3D Time-of-Flight image sensor features a high resolution three-tap pixel to capture more phases simultaneously, minimize motion artefacts, and optimize the use of light power – only one single train of pulses is required to capture all the phases. The sensor features spatial resolution of 832 x 600 pixels in both 2D and 3D as well as HDR capabilities for high flexibility of distance and reflectivity ranges.

Tichawa Vision GmbH


The BoroCIS camera is a solid-state solution for 100% high-speed inline scanning of the inner side of tubes, boreholes, and pipes. The camera scans the tube wall and converts the image to a rectangular frame without image distortion. A UV option can detect organic contamination, and a look-through option enables the device to inspect the tube shoulders as well as the tube walls.


xiX Detached Sensor Head Cameras

xiX cameras feature Sony Pregius S image sensors integrated into detached sensor heads, connected to the main camera electronics via flat ribbon micro-coax cable. This allows minimization of space requirements around the sensor while moving the rest of the camera out of the way. The distance can extend up to several meters. The camera also features PCIe interface with speeds of 32 or 64 gigabits per second.

Silver Honorees


Apera Vue

The Apera Vue computer vision solution for robot guidance uses pairs of 2D cameras and AI-powered stereo vision for 3D perception. It is designed to offer flexible range, field of view, and depth resolution. The AI engine can be trained to recognize any new object in just 24 to 48 hours using the object’s 3D model, without the need for image annotation.

AT - Automation Technology GmbH

3D Modular Compact Sensor

The 3D Modular Compact Sensor is based on a system of sensor, laser, and link modules for flexible scan width, measurement accuracy, measurement speed, triangulation distance, and working distance. Each sensor is individually assembled for optimal adaptation to application requirements. The sensor is also available as a dual camera version for multiple scan angles and doubling profile speed via alternating trigger mode.

Basler AG & CCS Inc.

Basler SLP Lighting Solutions

Basler SLP lighting solutions, developed in cooperation with CCS, are controlled using the pylon Camera Software Suite and designed to provide direct communication between lights and Basler ace U and ace L cameras. The components are automatically synchronized such that light strobing and overdrive functions are easy to access. The SLP lighting solutions are designed for ease and speed of setup and flexibility for applications with specific lighting requirements.

Computar Optics Group

MPT Series 45 Megapixel 1.4" Machine Vision Lenses

MPT Series 45 MPixel 1.4-inch C-Mount machine vision lenses are compact, lightweight, and engineered to optimize the capabilities of the latest industrial CMOS image sensors. The floating design provides performance at any working distance. The lenses feature high-level aberration correction and centering/alignment technology. The positioning of the high refractive index glass and low partial dispersion glass at crucial locations, the field curvature, and astigmatism are comprehensively corrected.

Computar Optics Group

LensConnect Remote Control Plug and Play Lens Series

LensConnect motorized machine vision C-Mount lenses enable remote, mobile lens management. The lens attains its power supply and communication control with one cable via USB interface. The setup software is compatible with Windows or Linux, and the operation program is open source. I2C interface is also available to allow camera manufacturers a direct connection to the lens.

DreamVu, Inc.

Omnidirectional 3D Vision Systems

The PAL omnidirectional 3D vision system uses a stereo lens based on an innovative mirror design and proprietary computational imaging techniques to create 360° RGB images and corresponding depth maps, that can be viewed as immersive 3D point clouds, using a single CMOS image sensor. It has no moving parts and lacks near-field blind spots by virtue of not depending on independent lenses.



CoaXPress-over-Fiber is an extension of the existing CoaXPress specification to support transport over fiber optics. Coaxlink QSFP+ frame grabber and Bridge IP Core exploits the capabilities of CoaXPress-over-Fiber. QSFP+ cards allow the equivalent of up to four CXP-12 connections per fiber optic cable and the Bridge IP cores to drive them, a solution towards higher speeds and deployment distances without compromising on synchronization.

iX Cameras, Inc.

i-SPEED 7 Series

i-SPEED 7 Series cameras feature proprietary Advanced Sensor Technology that produces CMOS sensors with increased light sensitivity, enhanced image clarity, and ultra-high-resolution at speeds up to 2.45 million frames per second. i-SPEED 7 SERIES image sensors have a 27.1 GPixel per second throughput rate, without any interpolation. The new sensor technology also offers proprietary black level control for deeper blacks and lower image noise.

Keyence Corporation of America

XT 3D Camera

The XT 3D camera uses four high-speed light projectors, each with red, green, and blue LEDs, to eliminate dead zones around the object being scanned and make sure all angles and features are covered by the projection. The camera also incorporates a large telecentric lens that allows for brighter images, decreased shutter speed, elimination of perspective error and distortion, and large field of view.

Laser Components USA, Inc.

Flexpoint MVpulse

FLEXPOINT MVpulse laser modules deliver power output between 10 megawatts and 100 megawatts and pulse lengths between 15 and 0.38 microseconds, depending on the specific model. The modules do not require extensive protective devices and feature a driver that ensures pulse lengths and frequency are correct within specified limits. The modules are offered in 640, 660, and 780 nm wavelengths.

Microchip Technology Inc.

Microchip Technology’s EQCO125X40 Family of CoaXPress 2.0 Devices

Microchip Technology’s EQCO125X40 CoaXPress 2.0 Devices seamlessly lock on all frequencies at all speed grades of the CoaXPress standard, from CXP-1 to CXP-12, eliminating the need for multiple channels by supporting 12.5 gigabits per second over a single cable. Integrated CDR improves jitter performance for the signal sent from the camera to the capture card. The devices can be used with cameras, frame grabbers, and active coax cables or links.

Mikrotron GmbH

EoSens Creation 2.0XGE

The EoSens Creation 2.0XGE smart camera combines high-speed imaging with in-camera image processing in an open platform. It can capture up to 2,240 frames per second at 1920 x 1080 resolution and process them with up to 5 gigabytes per second in real time. The camera features 10GigE interface, a XILINX Kintex UltraScale FPGA configured into a Mikrotron and custom IP block, and 2 GB of DDR4 memory reserved for custom IP.

Neurocle Inc.

Neuro-T, Deep Learning Vision Software with Auto DL

Neuro-T general-purpose deep learning software is designed for non-engineers to simplify the process of creating deep learning applications. The Auto Deep Learning Algorithm automates hyperparameter tuning, architecture selection, data augmentation, and model validation. The Data Management System with Server-Client Architecture allows running multiple devices off a single license and offers the multi-user friendliness of cloud-based software while protecting user data on a local server.

Photonfocus AG

MV4-D1280U-H01-GT UV Camera

The photonSPECTRAL UV camera combines sensitivity in the ultraviolet wavelength range with backside illumination and global shutter image sensor technologies. It offers a 1.3-megapixel CMOS image sensor that can capture 140 frames per second in 10-bit image output and has 40% quantum efficiency down to 170-nanometer wavelengths. The camera can also function as a UV linescan camera with speeds up to 70 kHz in 10-bit mode.

Pitcrew AI

Autonomous OTR (off the road) Tyre Inspection

Pitcrew AI is an automated inspection system for off the road tires, designed specifically for mining vehicles and to not require changes to site workflows. It is built around longwave infrared imaging and a multilayer R-CNN architecture, is self-contained in a solar-powered skid that can be deployed at the side of a road within a few hours, and detects tire separation and hot tires with detailed inspection reports.

Pleora Technologies

AI Gateway

The AI Gateway is a vision standards-based edge processing solution that enables a “hybrid AI” approach. It offers an integrated algorithm development platform, plug-in capabilities for common requirements and customization flexibility, and “no code” training. Designed to work with existing hardware and software, the platform integrates plug-in AI skills, a user-friendly approach to training, and an NVIDIA GPU to accelerate the development of advanced machine learning algorithms.


Metavision Intelligence Suite

The Metavision Intelligence Suite is a development environment for designing event-based vision systems. It offers 95 algorithms, 67 code samples, 11 ready-to-use applications, and modules for event-based vision training, prototyping, and development. The included toolkit allows developers to build event-based machine learning networks using an HD dataset. Open-source modules allow designers to build custom plugins that ensure compatibility with the Metavision Intelligence Suite.

Raptor Photonics Ltd.

Owl 640T

The Owl 640T camera is a size, weight, and power-optimized camera with InGaAs image sensor in 1/2-in. format featuring VGA resolution and 10 micrometer pixel pitch. It is sensitive in the 400 to 1700 nanometer range and features Automated Gain Control to enable best contrast between low light and bright, and a Non-Uniform Correction algorithm to correct for offset, gain, and dark current, performed by an on-board FPGA.

Saccade Vision


The Saccade-MD multi-directional and variable-density 3D scanning camera can locally optimize scanning positions within a large field of view to reduce overall image size by eliminating unneeded data. This provides higher scanning densities for specific features while maintaining a lower scan density in other areas. It combines a smart hybrid 3D sensor with flexible, autonomous software to provide the equivalent of 100 million 3D points per frame.

Teledyne e2v


Flash 4K and Flash 2K CMOS image sensors are specifically tailored for 3D laser triangulation. They can achieve 1800 frames per second at 4K resolution and 1500 frames per second at 2K resolution. The sensors include an up to 100 decibel real-time and single-frame HDR feature embedded on-chip for simultaneous inspection of high- and low-reflective materials, and ROI & Binning for frame rate optimization.

Teledyne e2v

Multi-focus Optical Module

The 2 megapixel multi-focus module offers an extended working range for industrial barcode readers and embedded systems for machine vision and Internet of Things devices. The module features MIPI interface and a 2 MPixel low-noise global shutter CMOS image sensor that leverages a novel mems-based reliable and rapid auto-focus technology provided by poLight ASA. The module provides fast image acquisition and a unique open-loop control and calibration technique.

Tichawa Vision GmbH


The RingCIS industrial scanner, based on proprietary Depth of Field CIS technology and available in monochrome or color models, provides distortion free images for 100% inline inspection of almost any extruded or pulled tube. The RingCIS can be inserted directly into the production process, immediately after extrusion, pulling, or welding, and provides a 360° gapless view along circumference and profile length.

TPL Vision

Modular Bar Light with Angle Changers

The Modular bar light with angle changers features white, red, or blue LED illumination and the ability to convert its lenses into narrow, medium, wide, or line light options using angle changers that employ an easy-to-use spring design for secure positioning. A black-out angle changer can effectively reduce the size of the bar. All of these features combined enable simplified feasibility testing for new applications.

Gold Honorees

Computar Optics Group

Computar ViSWIR Lens Hyper / Multi-Spectral Lens Series

Using ultra-low dispersion glass and low partial dispersion glass paired with superior design technology, the ViSWIR Hyper / Multi-Spectral Lens Series minimizes focus shift within a few microns in the 400 to 1700 nanometer visible and shortwave infrared range. A broadband anti-reflection coating applied to all lens elements minimizes adverse lighting effects. Spectral imaging is achievable with a single sensor camera simply by syncing the lighting without the need to adjust focus for different wavelengths.

Edmund Optics

TECHSPEC® LT Series Fixed Focal Length Lenses

TECHSPEC LT Series Fixed Focal Length Lenses combine an integrated liquid lens for quick autofocus, high resolution for 12-megapixel sensors up to 1.1-inch, and an easy-to-use interface requiring no mechanical adjustment. These lines allow machine vision system integrators to achieve extended depth of field while maintaining high throughput and speed. LT Series lenses can run quickly at F/#s down to f/2.8, have a response time in the order of < 2.5 ms and a total settling time of 15 ms, for a maximum 17.5 ms between focus locations.

Smart Vision Lights

Linear Tunable Field of View (LTF) Light Series

LTF, or linear tunable field of view lights, incorporate electronic controls to adjust light projection angles without any moving parts. The lights feature three-channel LEDs with 10-, 30-, and 50° lenses, with each channel having a dedicated lens angle. By setting the channels to different intensities, turning off individual channels, and tuning the field of view, the end user can produce unique beam-angle combinations. Specific configurations can be saved and returned to using the BTM-1000 Bluetooth accessory and SmartVisionLink app. The lights are available in standard wavelengths as well as infrared, shortwave infrared, and UV.


Zivid Two

The Zivid Two industrial 3D camera uses time-coded structured light to produce native color 3D point clouds. It is designed to image traditionally challenging shiny or reflective objects and has a 112 millimeter baseline for higher pixel coverage, especially relevant for imaging small parts. A unique combination of native colors and high dynamic range enable both single and mixed SKU bin applications and densely stacked or randomly arranged scenarios. Zivid Two’s new vision engine provides specific patterns, Artifact Reduction Technology, and new 3D hand-eye calibration and in-field correction tools. The camera is palm-sized, weighing less than 900 grams.

Platinum Honoree

Emergent Vision Technologies

Zenith HZ-100-G 100GigE Camera

It’s not a surprise that this year’s Platinum Honoree is a camera—the heart of any vision or imaging system. The Zenith HZ-100-G 100GigE Camera by Emergent Vision Technologies is this year’s platinum honoree.

The 103-megapixel HZ-100-G camera from the new Zenith series is based on the QSFP28-100GigE interface, providing 12,500 megabits per second bandwidth. The camera features the Gpixel GMAX32103 CMOS sensor, providing 30 frames per second in 8-bit modes at full 11276 x 9000 resolution.

QSFP28-100GigE provides low-cost commodity transceiver parts that allow a single mode fiber run of 10 km and beyond, usable in any speed application for easy forward and backward compatibility. The interface also offers low CPU overhead, low latency, low jitter, and accurate multi-camera synchronization using IEEE1588.

The camera is available in color and monochrome, is GigE Vision and Genicam compliant, has optional IP67 housing, and is appropriate for a wide range of applications including volumetric capture, flat panel inspection, and machine vision.

Voice Your Opinion

To join the conversation, and become an exclusive member of Vision Systems Design, create an account today!