Image Sensors World        Go to the original article...

AppleInsider quotes KGI analyst Ming-Chi Kuo on the oncoming iPhone "Face ID" design. It employs 4 cameras: a regular front camera, structured light 3D camera and proximity ToF sensor:


"According to Kuo, Apple's system relies on four main components: a structured light transmitter, structure light receiver, front camera and time of flight/proximity sensor.

Kuo points out that structured light transmitter and receiver setups have distance constraints. With an estimated 50 to 100 centimeter hard cap, Apple needs to include a proximity sensor capable of performing time of flight calculations. The analyst believes data from this specialized sensor will be employed to trigger user experience alerts. For example, a user might be informed that they are holding an iPhone too far or too close to their face for optimal 3D sensing."

Go to the original article...

Image Sensors World        Go to the original article...

PRNewswire: Inuitive introduces the NU4000, a multi-core vision processor that supports 3D Imaging, Deep Learning and Computer Vision processing for AR and VR, Drones, Robots and other applications. This next generation processor enables high quality depth sensing, "On-chip SLAM," Computer Vision and Deep Learning (CNN) capabilities.

NU4000 provides computing power exceeding a total of 8 Terra OPS, said to be the most powerful vision processor available:

• 3 Vector Cores that provide 500 Giga OPS
• A dedicated CNN processor that exceeds 2 Terra OPS enabling deep neural networks such as VGG16 reaching 40 frames (ROIs) per second at 10 times less power of the equivalent GPU, DSP or FPGA implementations
• 3 Powerful CPU Cores that provide more than 13,000 CoreMark
• Depth processing engine that delivers a throughput of 120Mp/s and supports multiple simultaneous streams of stereo and structured light
• SLAM engine enabling accurate key point extraction at 120fps from 2 cameras simultaneously
• Advanced Time-Warp HW engine that reduces the Motion-to-Photon latency to 1msec for extensive VR and MR use cases
• More than 3MB of on-chip SLAM servicing the vision cores
• High throughput LPDDR4 interface that reduce external memory access bottlenecks
• Connects to 6 cameras and 2 displays
• Chip area 7 x 8mm2 in 12nm process

Go to the original article...

Image Sensors World        Go to the original article...

DeviceSpecifications: These days, dual rear camera is used even in very low end phones, such as this one:

Go to the original article...

Image Sensors World        Go to the original article...

TU Delft and Harvest Imaging publish an open access paper "Temperature Sensors Integrated into a CMOS Image Sensor" by Accel Abarca, Shuang Xie, Jules Markenhof, and Albert Theuwissen. Apparently, the paper is a continuation of the MSc thesis "Integrating a Temperature Sensor into a CMOS Image Sensor."

"The test image sensor consists of pixels and temperature sensors pixels (=Tixels). The size of the Tixels is 11 μm × 11 μm. Pixels and Tixels are placed next to each other in the active imaging array and use the same readout circuits. The design and the first measurements of the combined image-temperature sensor are presented."

Go to the original article...

Image Sensors World        Go to the original article...

Researches from University of Utah at Salt Lake City proved that AI can be somewhat successful in distinguishing between the digits as seen by image sensor with no lens. "Lensless-camera based machine learning for image classification" paper by Ganghun Kim, Stefan Kapetanovic, Rachael Palmer, and Rajesh Menon is published by arxiv.org. From the abstract:

"Finally, we demonstrated that the trained ML algorithm is able to classify the digits with accuracy as high as 99% for 2 digits. Our approach clearly demonstrates the potential for non-human cameras in machine-based decision-making scenarios."

Go to the original article...

Image Sensors World        Go to the original article...

SEMI European Imaging & Sensors Summit 2017 is to be held on September 20-22 in Grenoble, France. The summit agenda has many interesting presentations:

• Sensor as a solution
  Chae Lee, Senior VP, LG Electonics
• In the era of mixed reality and augmented environment sensing, what does innovation mean? What is Leti’s vision?
  Marie Noelle Semeria, CEO, Cea-Leti
• 3D Time-of-Flight Cameras in Industrial Applications
  Thomas Kuhnke, Senior Electronics Design Engineer, 3D Business, Basler AG
• Hybrid semiconductor direct conversion CMOS x-ray imaging detectors with application examples
  Tuomas Pantsar, Chief Technology Officer / Charge Integration, Oy Ajat
• CMOS based microdisplays, imager and sensors enhanced by OLED/OPD integration
  Philipp Wartenberg, IC Design Engineer and Project Manager / Deputy head of department IC and System Design, Fraunhofer Institute for Organic Electronics, Electron Beam and Plasma Technology FEP
• TSMC Imaging Technology to Enable Innovation in AR/VR Applications
  Tripti Bhanti, Senior Technical Manager, TSMC Europe BV
• Fusion Bonding Enabler for Backside Illuminated Image Sensors - What’s Next
  Thomas Uhrmann, Director of Business Development, EVG
• Technological Trends in Thermal Image Sensors
  Christel-loic Tisse, Technical & Innovation Director, ULIS
• Packaging of Image Sensors and Devices
  Lutz Mattheier, Manager Assembly Technology Development, First Sensor Microelectronic Packaging GmbH
• Image Fusion: How to Best Utilize Dual Cameras
  Roy Fridman, Director of Product Marketing, Corephotonics
• High Performance CMOS Integrated Graphene Photodetectors
  Tapani Ryhänen, CEO, Emberion
• CMOS image sensor for bio-medical and scientific applications
  Renato Turchetta, CEO, IMASENIC
• New developments in logarithmic pixels and sensors
  Yang Ni, Founder & CTO, NIT
• Colour X-ray Imaging Solutions for Non-Destructive Testing based on Photon Counting Technology
  Juha Kalliopuska, Chief Executive Officer and Co-founder, ADVACAM
• High resolution global shutter image sensors for machine vision and 8K video
  Guy Meynants, Engineering Fellow, ams/CMOSIS
• VTT’s hyperspectral imaging technology - from high-performance applications towards volume scalability
  Anna Rissanen, Research Team Leader, VTT
• Image Sensors for Future VR
  Yiwan Wong, Partnership Lead, Oculus Research
• Ambient Sensing – New ways how intelligent devices can cope with the environment
  Roland Helm, Segment Head Sensors, Infineon Technologies AG

Go to the original article...

Image Sensors World        Go to the original article...

Leti announces readiness of the new technology of very high-resolution (>1000 dpi) fingerprint sensor based on a matrix of interconnected piezoelectric ZnO nanowires:

Go to the original article...

Image Sensors World        Go to the original article...

Sony announces Venice - the next generation digital cinema camera featuring 4K full-frame sensor capable of 60fps or 6K at 30fps speed. Venice is said to have a high speed readout which minimizes the jello effects typical in the CMOS sensors. The new sensor DR is promised to achieve 15+ stops.

Go to the original article...

Image Sensors World        Go to the original article...

Recently, automotive LiDAR startups attract a lot of VC money:

LeddarTech announces a combined investment of US$101m led by Osram and including Delphi, Magneti Marelli and Integrated Device Technology as strategic investors, as well as Fonds de solidarité FTQ. This round of funding will allow LeddarTech to enhance its ASIC development efforts, expand its R&D team, and accelerate ongoing LiDAR development programs with select Tier-1 automotive customers for rapid market deployment.

“We are thrilled to welcome such an influential group of companies as strategic investors, with whom we share a common goal of making mobility safer and more efficient,” stated Charles Boulanger, LeddarTech’s CEO. “Their strong endorsement is a testament to the maturity of the Leddar technology and the credibility of our product roadmap to support the ongoing development of autonomous driving. This positions LeddarTech as the definitive reference in solid-state LiDAR and validates our Tier-1 partnership business model to rapidly deploy LiDARs on a commercial scale in automotive applications. We believe this announcement represents a major stepping stone towards achieving market-share leadership in the multibillion-dollar automotive LiDAR market by 2020.”


PRNewswire: Innoviz has raised $65m in Series B funding as its LiDAR moves into mass production. Delphi Automotive PLC and Magna International participated in the round, along with additional new investors including 360 Capital Partners, Glory Ventures, Naver and others. All Series A investors, including Zohar Zisapel, Vertex Venture Capital, Magma Venture Partners, Amiti Ventures and Delek Motors, participated in this round. A second closing of this round is expected to be announced soon, introducing additional strategic partners.

"As the autonomous driving market matures, Innoviz has clearly established itself among industry leaders not only as the best-in-class LiDAR solution but also as an integral part of the overall autonomous vehicle stack," said Omer Keilaf, Co-founder and CEO of Innoviz. "Today's financing from strategic partners such as Magna and Delphi demonstrates their support for Innoviz as the industry's leading option for high-performing LiDAR and is further proof that we have moved into the next phase of our growth. Given both Magna's and Delphi's unwavering commitment to high performance and high safety standards, investing in and partnering with Innoviz is a natural choice. With their joint efforts, we will be able to scale more quickly and put autonomous driving technology on the road much faster."

Innoviz technology that leverages the company's proprietary System, MEMS and Detector designs to give autonomous vehicles sensing capabilities even in challenging environments such as bright direct sunlight, varying weather conditions and multi-LiDAR environments.



Oryx Vision announces a $50m Series B funding round. Third Point Ventures and WRV led the round, which was joined by Union Tech Ventures, and existing investors Bessemer Venture Partners, Maniv Mobility and Trucks VC. A mere 15 months after its first funding round, this fundraise brings the total investment in Oryx to $67M.

Oryx will use the new funds to accelerate its development activities and to intensify its commercial engagements with car OEMs, tier-1 supplier and technology players. Having demonstrated the unique capabilities of its technology over the past year, the company expects to ship units for car-mounted testing in the second half of 2018.

“Oryx is building the first solution that will meet all the key requirements of automotive LiDARS – high performance, car durability and low price – without a tradeoff. We are delighted to receive a vote of confidence in our vision from such sophisticated investors, and to have the resources to bring this technology to market quickly and at the highest quality,” says Rani Wellingstein, Oryx co-founder and CEO.

From Oryx Vision presentation:

Go to the original article...

Image Sensors World        Go to the original article...

Aphesa publishes a blog post "Why You Can't Use A White (Broad Spectrum) Light Source For Your Spectral Measurements." Few quotes:

"The R, G, B or the monochrome curve is not the response curve of the sensor but only some information about it.

First of all, it is usually only measured for an incident beam perpendicular to the sensor, or in other words at a zero chief ray angle. But we know that the spectral response varies in intensity and in shape with the angle of incidence.

Secondly, it is a noisy measurement and therefore the precision depends on the number of samples taken. The measurement is noisy because the amount of light after the monochromator is limited as only a tiny fraction of the source's spectrum reaches the sensor. Therefore the signal level is small and its SNR is therefore low.

The shape of the spectrum also depends on the bandwidth of the measuring instrument. As the response spectrum can exhibit oscillations at several scales (see this other publication about spectral response), only a very small bandwidth, i.e. a narrow monochromator lid, can reproduce the actual shape of the response curve, any other approach will only produce a smoothed curve without any detail.

Finally, the wide band light source will change over time and its spectrum will change over temperature, therefore requiring regular calibration.

The light source that we use is based on a femtosecond laser and a supercontinuum.

The femtosecond laser is a solid laser that provided very high power light pulses with a duration in the range of the femtosecond. Lasers are known to provide coherent, repeatable and high intensity light. Our laser is red and the laser has the size of a small table.

The role of the supercontinuum is to turn the monochromatic laser light into a wide spectrum by a collection of non-linear processes. The supercontinuum is a long microstructured optical fiber. The fiber seems red at its beginning and is white in the end as the spectrum broadens along the fiber."

Go to the original article...

Image Sensors World        Go to the original article...

Dave Etchells from Imaging Resource publishes a nice overview of Sony image sensor fab. It describes a process of growing the silicon boules by Czochralski method in great details, then gets sketchy on photolitography and other FEOL and BEOL steps, then again goes into much more details on dicing, wire bonding, and packaging. While very simplistic, the article describes quite a complete process flow at Sony Kumamoto fab.

Go to the original article...

Image Sensors World        Go to the original article...

Espoo-Finland and Cambridge, UK-based startup Emberion is a spin-off of Nokia. The new company works to commercialize graphene-based microbolometric image sensors:

Go to the original article...

Image Sensors World        Go to the original article...

Fayçal Saffih from University of Waterloo, Canada publishes his video lecture on "Implementations of Intelligence on Smart CMOS Image Sensors" given on Sept. 1, 2017 in York University:

Go to the original article...

Image Sensors World        Go to the original article...

Huawei unveils Kirin 970 mobile SoC "combining the power of the cloud with the speed and responsiveness of native AI processing."

"Mobile AI = On-Device AI + Cloud AI. HUAWEI is committed to developing smart devices into intelligent devices by building end-to-end capabilities that support coordinated development of chips, devices, and the cloud. The Kirin 970 is the first in a series of new advances that will bring powerful AI features to our devices and take them beyond the competition," says Richard Yu, CEO of Huawei Consumer Business Group.

Go to the original article...

Image Sensors World        Go to the original article...

Autosens publishes ST presentation on its HDR ISP for automotive cameras from the converence in Detroit in May 2017:

Go to the original article...

Image Sensors World        Go to the original article...

Teledyne DALSA demonstrates its Wafer-Level-Packaged (WLP) 320 x 240 Vanadium Oxide microbolometer. Why is it such a big deal? The vacuum packaging cost is a major contributor to the micorbolometer imager price and its reduction is a breakthrough for the whole industry.

The company says: "Our novel wafer-level packaging (WLP) approach to microbolometer manufacturing in an optimized MEMS infrastructure gives us the ability to dramatically alter the traditional price-performance tradeoff.

Our advanced capabilities in wafer processing allows us to bond multiple 200 mm wafers precisely in a 3D stack that eliminates the need for external chip packaging--packaging which can account for 75% or more of device cost.

By stacking a cap wafer, using world-leading vacuum sealing technology, onto a CMOS high-speed readout circuit (ROIC) wafer populated with VOx pixels, we deliver smaller, lighter devices with leading edge performance and game-changing lower costs, enabling compact detectors with advanced integrated features that will drive the next phase in the evolution of uncooled IR imaging."

Go to the original article...

Image Sensors World        Go to the original article...

Photonics.com publishes Teledyne DALSA article "Sensor Market Set to Soar." It nicely re-arranges the IC Insights data to show the growth potential of different markets:

Go to the original article...

Image Sensors World        Go to the original article...

Renesas has been formed from combined semiconductor businesses of Mitsubishi, Hitachi and NEC, all of them used to be image sensor manufacturers in the past. However, up to now, Renesas was not in image sensor business. This was especially surprising as the company kept filing for new image sensor patents and publishing papers over all these years while having no image sensing products in its portfolio. Now, this paradox has been resolved...

BusinessWire: Renesas announces a new high-sensitivity 8.48MP CMOS sensor (RAA462113FYL) for 4K security cameras. The company also talks about existing 2.12MP product that is already in mass production, but has never been widely announced.

The new sensor supports a line-by-line HDR mode in which long exposure data and short exposure data are output separately for each line. This allows video capture for high-contrast scenes. Samples of the RAA462113FYL are now available and mass production is scheduled to begin in December 2017.

Go to the original article...

Image Sensors World        Go to the original article...

EETimes publishes Corephotonics article "Evolution of Zoom Camera in Smartphones." Few quotes:

"The future of mobile dual cameras is brighter than ever. Adoption rates surpassed analysts' predictions, and all major OEMs plan to launch or have already launched handsets with zoom dual cameras. Now the race begins for higher zoom factors, better optical stabilization, lower Z-heights and better low-light performance — anything that improves the overall camera and handset user experience. In this paper, we discussed some ways to increase the zoom factor and reduce the Z-height while maintaining superior image quality."

Oppo camera with 5x folded zoom and 2 axes OIS is said to represent the latest advances in smartphone zooms:

Go to the original article...

Image Sensors World        Go to the original article...

Autosens publishes videos from its Detroit conference in May 2017. Here is few of them talking about image sensing:

Go to the original article...

Image Sensors World        Go to the original article...

PRNewswire, GlobeNewsWire: Qualcomm and Himax jointly announce a collaboration on low power active 3D depth sensing camera for use cases such as biometric face authentication, 3D reconstruction, and scene perception for mobile, IoT, surveillance, automotive and AR/VR.

Qualcomm expertise in computer vision and algorithm with Himax's technologies in wafer optics, sensing, driver, and module integration are combined in a fully integrated SLiM (Structured Light Module) 3D solution. The SLiM is a turn-key 3D camera module that delivers real-time depth sensing and 3D point cloud generation with high resolution and high accuracy for indoor and outdoor environments. Qualcomm and Himax will commercialize the SLiM for a wide array of markets and industries with mass production targeting in Q1/2018.

"Our 3D sensing solution will be a game changing technology for smartphones, where we will enable the Android ecosystem to provide the next generation of mobile user experience," said Jordan Wu, President and CEO of Himax. "Our two companies have worked together for more than four years to design the SLiM 3D sensing solution to meet growing demands for enhanced computer vision capabilities that will enable amazing new features and use cases in a broad range of markets and applications. We are pleased to partner with Qualcomm Technologies to put together an ecosystem and to enable the revolutionary computer vision solutions for our customers globally in a timely fashion."

Go to the original article...

Image Sensors World        Go to the original article...

Macrumors publishes images of Lexus car with Apple-equipped set of LiDARs and cameras. Apparently, Apple is testing different configurations and comparing the data quality coming from them:


Thanks to OB for the link!

Go to the original article...

Image Sensors World        Go to the original article...

Yole Developpement releases "Automotive sensing: a mature yet highly dynamic market" report. Image sensors of all sorts is a significant part of this market:

"Despite just 3% growth in the volume of cars sold expected through to 2022, Yole expects an average growth rate in sensors sales volumes above 8% over the next five years, and above 14% growth in sales value. This is thanks to the expanding integration of high value sensing modules like RADAR, imaging and LiDAR.

Among all sensing technologies located in the car, three main sensors will drastically change the landscape: imaging, RADAR and LiDAR sensors. Imaging sensors were initially mounted for ADAS purposes in high-end vehicles, with deep learning image analysis techniques promoting early adoption. It is now a well-established fact that vision-based AEB is possible and saves lives. Adoption of forward ADAS cameras will therefore accelerate.

Growth of imaging for automotive is also being fueled by the park assist application, and 360° surround view camera volumes are skyrocketing. While it is becoming mandatory in the US to have a rear view camera, that uptake is dwarfed by 360° surround view cameras, which enable a “bird’s eye view” perspective. This trend is most beneficial to companies like Omnivision at sensor level and Panasonic and Valeo, which have become the main manufacturers of automotive cameras.

LiDAR remains the “Holy Grail” for most automotive players, allowing 3D sensing of the environment. In this report Yole’ analysts highlight the different potential usages of this technology, which will transform the transportation industry completely.

“We expect tremendous growth of the LiDAR market within the next five years, from being worth US$300 million in 2017 to US$4.4 billion by 2022,” details Guillaume Girardin from Yole. LiDAR is expected to be a key technology, but sensing redundancy will still be the backbone of the automotive world where security remains the golden rule."

Go to the original article...

Image Sensors World        Go to the original article...

TomsHardware: Intel Movidius releases Miriad X Vision Processing Unit (VPU) with a dedicated Neural Compute Engine. The SoC includes vision accelerators, a Neural Compute Engine, imaging accelerators, and 16 SHAVE vector processors with a total of up to 4 TOPS (Trillion Operations Per Second) of performance in 1.5W power package:


Medium, China Money Network: China-based Cambricon AI startup raises $100M at reported $1b valuation. Cambricon was founded by Prof. Chen Tianshi in 2016 as a spin-off from the Institute of Computing Technology of the Chinese Academy of Science. Its processor, the Cambricon-1A, is claimed to be the first commercial chip for deep learning applications and can be used in robotics, drones, autonomous vehicles and consumer electronics. It is said to have a better performance when running mainstream AI algorithms, is more energy efficient and has a higher integration density.

The total revenue of the AI-related deep learning chip market is forecast to rise from $500M in 2016 to $12.2b in 2025 at CAGR of over 40%, according to Tractica market research.


Update: Digitimes reports that Cambricon expects to roll out hundreds of millions of AI SoCs for smart terminal devices and servers in the next three years, the CEO Chen Tianshi said. Cambricon will license its processor instruction set to more enterprises engaged in AI applications. In this regard, Huawei's HiSilicon Kirin 970 chipset, to be released in September, is to carry the IP instruction set for Cambricon-1A co-processor.

Go to the original article...