Monday, September 30, 2019

Local TV Reports on ON Semi Rochester Layoffs

Rochester, NY, local TV station reports that ON Semi announced layoffs of 30 of its employees:

"A company spokesperson says it's shutting down space used to create certain image sensors, which are no longer in high demand in the tech industry.

“ON Semiconductor is committed to staying in in the New York area through recently announced partnerships,” said Michael Miller, general manager and director of operations at ON Semiconductor. “Our Rochester site will still support pixel design and modeling, device characterization and test engineering. In additional to these engineering functions, certain marketing functions for our Intelligent Sensing Group will remain onsite.”

Thanks to DG for the link!

Sony Reorgs IS Group

Correction: The IS is Information Services Division. This news does not have any relation to image sensor business.

Sony announces organizational and management changes in image sensor information services business. It appears that image sensors moved from Imaging & Sensing Solutions to Corporate IS Department. Also, the matrix R&D structure of Development Division #1 and #2 has been morphed into Application Technology Development, System and Platform Technology Development, and Fundamental Technology Research and Development organizations. No explanation for the re-org is given.

Sunday, September 29, 2019

JP Morgan Forecasts 320Munits 3D Camera Sales in 2020

IFNews quotes JP Morgan report forecasting 320M 3D sensing cameras in smartphones in 2020, 260M of them - in Apple iPhones:

Saturday, September 28, 2019

Automotive News from Autosens Brussels

EETimes reporter Junko Yoshida publishes "ADAS: Key Trends on ‘Perception’" from Autosens Brussels held in mid-September 2019:
  • At this year’s AutoSens in Brussels, assisted driving (ADAS), rather than autonomous vehicles (AV), was in sharper focus. Clearly the engineering community is no longer in denial. Many acknowledge that a big gap exists between what’s possible today and the eventual launch of commercial artificial intelligence-driven autonomous vehicles.
  • Many vendors are adding more intelligence at the sensory node, by fusing different sensory data (RGB camera + NIR; RGB + SWIR; RGB + lidar; RGB + radar) right on the edge.
  • Trieye claims it has found a way to design SWIR by using CMOS process technology. “That’s the breakthrough we made. Just like semiconductors, we are using CMOS for the high-volume manufacturing of SWIR cameras from Day 1,” said Avi Bakal, CEO and co-founder of Trieye.
  • EE Times’ subsequent discussion with Trieye revealed that Outsight will be using Trieye’s SWIR camera. Outsight is promoting its sensor fusion box, scheduled for sampling in the first quarter of 2020.
  • AEye is finishing the combined RGB and lidar system, automotive grade, embedded with AI in the next three to six months. The price will be “less than $1,000.”

Friday, September 27, 2019

Yole on Automotive Vision Trends

Yole Developpement publishes a webinar "Automotive Imaging on its Way to Sensing" with cars cameras multiplications starting to look quite similar to the smartphone market few years ago:

ISP Optimization for Automotive Applications

MDPI publishes Valeo paper "Overview and Empirical Analysis of ISP Parameter Tuning for Visual Perception in Autonomous Driving" by Lucie Yahiaoui, Jonathan Horgan, Brian Deegan, Senthil Yogamani, Ciarán Hughes, and Patrick Denny.

"Image quality is a well understood concept for human viewing applications, particularly in the multimedia space, but increasingly in an automotive context as well. The rise in prominence of autonomous driving and computer vision brings to the fore research in the area of the impact of image quality in camera perception for tasks such as recognition, localization and reconstruction. While the definition of “image quality” for computer vision may be ill-defined, what is clear is that the configuration of the image signal processing pipeline is the key factor in controlling the image quality for computer vision. This paper is partly review and partly positional with demonstration of several preliminary results promising for future research. As such, we give an overview of what is an Image Signal Processor (ISP) pipeline, describe some typical automotive computer vision problems, and give a brief introduction to the impact of image signal processing parameters on the performance of computer vision, via some empirical results. This paper provides a discussion on the merits of automatically tuning the ISP parameters using computer vision performance indicators as a cost metric, and thus bypassing the need to explicitly define what “image quality” means for computer vision."

Alexis Lluis Gomez, Senior Manager of Image Quality, ARM, talks about automotive ISP optimization too in his Autosens Detroit presentation:

Manjunath Somayaji, Director of Imaging R&D, GEO Semiconductor, presents its view on ISP optimization:

On its side, Algolux says that ISP should be tightly integrated into the whole CV pipeline, rather than optimized as a separate block:

Thursday, September 26, 2019

TechInsights Analyses iPhone 11 Cameras

TechInsights publishes a teardown report of the new Apple iPhone 11 Pro Max smartphone with some info on its cameras:

"We were excited during the September Apple Special Event to hear of 100% Focus Pixels for iPhone’s wide-angle camera. We expected this to coincide with full-chip dual photodiode phase detection autofocus (PDAF), however our initial findings show a familiar PDAF pattern comparable to that used for 2018 iPhone’s wide-angle camera. The analysis in progress will determine if Apple has indeed gone to dual PD as 100% Focus Pixels suggests. If true, it would be the first use of masked + dual PD PDAF we have documented."

MIPI Alliance Announces CSI‑2 v3.0, C-PHY v2.0, and D-PHY v2.5

BusinessWire: MIPI Alliance announces enhancements to MIPI Camera Serial Interface-2 (MIPI CSI‑2), the most widely used camera specification in mobile and other markets.

MIPI CSI‑2 v3.0 introduces three key new features:
  • Unified Serial Link (USL), for encapsulating connections between an image sensor module and application processor—This capability is crucial for reducing the number of wires needed in IoT, automotive and client products for productivity and content creation, such as all-in-one and notebook computers.
  • Smart Region of Interest (SROI), for analyzing images, inferencing algorithms and making better deductions—This capability could enable machines on a factory floor, for example, to more quickly identify potential defects on a conveyor belt, or it could enable medical devices to more surely recognize anomalies such as tumors.
  • RAW-24, for representing individual image pixels with 24-bit precision—This capability will enable machines to make decisions from superior quality images. An autonomous vehicle, for example, could decipher whether darkness on an image is a harmless shadow or a pothole in the roadway to be avoided.

Innovation of MIPI CSI‑2 never stops; we aim to remain at the frontier of providing end-to-end imaging conduit solutions for emerging vision and real-time perception and decision making AI applications mapped to mobile, client, IoT, medical, drones and automotive (ADAS) product platforms,” said Haran Thanigasalam, MIPI Camera Working Group chair. “In fact, work is already well underway on the next version of MIPI CSI‑2, with a highly optimized ultra-low-power always-on sentinel conduit solution for enhanced machine awareness, data protection provisions for security, and functional safety, as well as MIPI A-PHY, a forthcoming longer reach physical layer specification.

MIPI C-PHY v2.0 was recently released to support CSI‑2 v3.0 capabilities, including support for 6 Gsps over the standard channel and up to 8 Gsps over the short channel; RX equalization; fast BTA; medium channel lengths for IoT applications; and an in-band control signaling option. MIPI D-PHY v2.5, with alternate low power (ALP), which uses pure low-voltage signaling instead of legacy 1.2 V LP signaling and a fast BTA feature for support of CSI‑2 v3.0, will be released later this year.

ams Launches Extended Range dToF Module

Businesswire: ams announces the world’s smallest integrated module for direct ToF distance measurement in range from 2cm up to 2.5m. The TMF8801 is said to be more than 30% smaller than competing ToF sensors yet offers good accuracy and usability in the presence of sunlight. Competing ToF sensors struggle to accurately measure distance below dirty or smudged cover glass whereas the TMF8801 maintains high accuracy through use of its on-chip histogram processing.

The new TMF8801 direct ToF sensor which complements the TMF8701 short range ToF sensor released in 2018, provides extended distance range sensing for smartphones,” says Dave Moon, Senior Product Marketing Manager in the Integrated Optical Sensors business line at ams. “The TMF8801 is ideally suited to provide support for improving world-facing mobile phone camera performance using laser detection auto-focus (LDAF), enabling mobile phone users to take pin-sharp photos and selfies.

With an integrated 940nm VCSEL, multiple SPADs, time-to-digital converter, and on-chip microcontroller for processing histograms, the TMF8801 delivers superior ToF performance. Compared to distance averaging employed with an indirect ToF system, the direct ToF time measurement methodology used in the TMF8801 is said to deliver higher accuracy distance measurements. When used to support LDAF, high accuracy distance measurements deliver improved camera performance.

The small size and low power consumption of the TMF8801 is suitable for collision avoidance detection used in autonomous vacuum cleaners, and industrial robotics. For display-based products such as mobile computing, the TMF8801 can be used for user presence detection to automatically wake up or put the system into a low-power sleep mode based on the presence or absence of a user.

The TMF8801 is housed in a 2.2mm x 3.6mm x 1.0mm package and is in volume production. Unit pricing is $2.89 in order quantities of 5,000 pieces.

The previous generation TMF8701 sensor has a shorter range up to 60cm in a similarly sized package. It has won an number of awards.

Sony Unveils 47MP MFT Sensor for Security and Industrial Applications

Sony IMX492LLJ/LQJ 1.4-inch sensor with 2.315um pixels has quite high readout speed:

Panasonic Talks about 8K Organic Sensor inside Broadcast TV Camera

DigitalCameraWorld, YMCinema: Panasonic publishes an interview on its prototype of 8K broadcast camera based on the organic film image sensor:

Thanks to TG for the link!

Wednesday, September 25, 2019

TowerJazz Announces the Availability of its Stacked BSI Hybrid Bonding Technology

GlobeNewswire: TowerJazz releases its advanced 65nm stacked BSI hybrid bonding technology manufactured in Company’s 300mm facility in Uozu, Japan. This new offering, implemented in pixel level, enables a reduced pixel size along with enhanced performance by splitting the pixel IC functionality between two wafers. With a pitch smaller than 4µm, sensors manufactured with this technology provide higher data rate, lower power consumption, better performance, and smaller form factor, accurately meeting the innovative requirements of world-leading growing markets and applications including automotive, high-end photography, event-driven sensors and ToF, among others.

We are very pleased to announce this new and highly advanced offering, providing our customers with extensive technological solutions which enable them to competitively lead in their markets,” said Avi Strum, SVP and GM of the Sensor Business Unit at TowerJazz. “As always, we are greatly committed to bring to market the most advanced technology and state-of-the-art pixel performance providing strategic, value-add development opportunities and roadmap.

With pixel functionality split between wafers, TowerJazz hybrid BSI stacking, electrically connects each of the PD in the imaging top wafer to the pixel circuit in the bottom wafer. This new offering is an expansion of the Company’s CIS technology which includes stitching with 1D and 2D for large format sensors in the high-end photography, industrial and medical X-Ray markets, as well as advanced SPAD technology profoundly used in the rapidly growing automotive market with silicon proven pixels that can be customized to customer specifications.

Prophesee Introduces Event-Based Sensor in Industry-Standard, Cost-Efficient Package

PRNewswire, EETimes: Prophesee introduces what it calls the first industry-standard packaged chip with Event-Based Vision technology, a significant advancement over traditional frame-based vision approaches.

Under development for five years and commercially tested, this third-generation version is a VGA sensor with 15 μm pixels in a 3/4" optical format. It is aimed at developers of cameras in industrial automation and IoT systems such as robots, inspection equipment, monitoring and surveillance devices. It leverages Prophesee's neuromorphic vision technology to offer efficient capabilities for a variety of use models, including ultra high-speed part counting, vibration measurement and monitoring or kinematic monitoring for predictive maintenance.

"This is a major milestone for Prophesee and underscores the progress in commercializing our pioneering Event-Based Vision sensing technology. After several years of testing and prototyping, we can now offer product developers an off-the-shelf means to take advantage of the benefits of our machine vision inventions that move the industry out of the traditional frame-based paradigm for image capture," said Luca Verre, co-founder and CEO of Prophesee.

In Prophesee Metavision sensor, each pixel is independent and asynchronous, only activating if it senses a change in the scene, a movement - an event. This proprietary Event-Based approach allows for major reductions of power, latency and data processing requirements imposed by traditional frame-based systems. It enables sensors to achieve much higher DR than commonly associated with high-speed vision. And it allows cost-efficient sensors and systems to record events that would otherwise require conventional cameras to run at 10,000 fps.

The chip is packaged in a 13x15 mm mini PBGA package and is manufactured in a 0.18um CIS process. The packaging was realized by Kingpak.

ON Semi-Rochester Stops Making CCDs

ON Semi announces that it discontinues all its CCD products (also here.) The rumor says that the last reminder about Kodak image sensor business, the Rochester, NY, CCD fab will be closed on June 20, 2020 (confirmed from 3 independent sources.)

"The ongoing market conversion from CCD to CMOS image sensors has been accelerated both by an industry slowdown that began in 2018 as well as availability of advanced CMOS image sensors such as ON Semiconductor’s line of XGS products. In light of these factors, ON Semiconductor has been monitoring the sales activity and viability of our CCD image sensor products. The recent activity for the products on the attached device list has declined to a level which can no longer be sustained. As a result of this analysis we regret to announce discontinuance of our CCD products as described in this notification.

All CCD products are impacted by this change. The list of OPNs provided with this notification includes suggested replacement parts (where it was possible to identify a suitable replacement). In many cases, a call to ON Semiconductor Sales may be appropriate to discuss XGS CMOS image sensors in detail. For those devices that have no recommended replacement identified, ON Semiconductor has partnered with Rochester Electronics who will carry any remaining device inventory to potentially allow for a continued source after ON Semiconductor has completed its business process.

PLEASE NOTE: ON Semiconductor may implement minimum order quantities or dollar amounts based on inventories available. Also, Devices on this Product Discontinuance become NCNR (Non-Cancelable, Non-Returnable).

SiOnyx vs Hamamatsu Patent Litigation Wrap-Up

Fish and Richardson Litigation Blog publishes a summary of SiOnyx vs Hamamatsu Photonics (HPK) patent lawsuit:

"At trial, the jury found that the HPK willfully infringed SiOnyx’s U.S. Patent No. 8,080,467. The jury also found that HPK breached a non-disclosure agreement it had with SiOnyx, because “HPK took [SiOnyx’s] inventive ideas, added other inventive ideas to it, and obtained [nine] patents as a result.”

The jury awarded SiOnyx $796,469 in damages for breach of contract and $580,640 in damages for unjust enrichment. But, despite finding willful patent infringement, the jury awarded $0 in damages for patent infringement.

SiOnyx moved for “an injunction awarding [it] ownership of [HPK’s nine] patents….” The Court considered the jury’s finding that “HPK [] took confidential information SiOnyx had entrusted to it and used it for its own gain [by filing nine patents].” ....The Court thus granted an affirmative injunction “awarding and transferring ownership of [HPK’s nine] patents to SiOnyx[].”

The Court explained that “SiOnyx is a tiny company relying on a single type of technology, whereas HPK is a large company with a broad and diversified range of products.” Thus, the Court granted an injunction prohibiting HPK from “making, using, offering for sale, selling, or importing products infringing [SiOnyx’s] ’467 patent.”

HPK appealed this case to the Federal Circuit in late August, and that appeal is pending.

Tuesday, September 24, 2019

GPixel Announces Cost Effective 4K Dual Line Scan Sensor

Gpixel announces the GL0402, a 4096 x 2 resolution, 7um square pixel size, global shutter line scan image sensor with 200kHz line rate offering a cost effective solution for various in line inspection and sorting applications.

GL0402 offers a read noise of less than 5e- and full well charge of higher than 11k e-, resulting in 66dB intra-scene DR. DR and sensitivity can be further increased by utilizing on-chip 1x2 and 2x2 charge domain binning. Maximum line rate of 200kHz line rate in single line and 100kHz line rate in dual-line mode is achieved over 12 pairs of sub-LVDS high speed outputs allowing data rates up to 10.9Gbit/sec.

We are pleased with this new GL0402 line scan sensor enabling our customers to develop next generation, cost effective in-line inspection systems,” commented Wim Wuyts, Chief Commercial Officer of Gpixel. “This new line scan product release also confirms our continued focus at line scan sensors developments which is an essential part of our fast-growing industrial inspection product portfolio.

The sensor is assembled in a compact 76 pins CLCC and is offered in two chroma versions: the monochromatic sensor and RGB Bayer color. Samples are available by September 2019.

Samsung Unveils Industry's First 0.7um Pixel

BusinessWire: Samsung introduces the industry’s first 0.7μm pixel sensor, the 43.7MP ISOCELL Slim GH1.

Samsung has been stepping up in pixel technology innovation from the industry’s first 1.0μm-pixel image sensor, to most recently, 0.8μm ultra-high-resolution sensors at 64Mp and 108Mp,” said Yongin Park, EVP of the sensor business at Samsung Electronics. “We are pleased to deliver yet another breakthrough with the industry’s first 0.7μm pixel image sensor, the ISOCELL Slim GH1 that will enable sleeker and more streamlined designs as well as excellent imaging experiences in tomorrow’s smartphones.

ISOCELL Plus, Samsung’s latest pixel isolation technology, minimizes color cross-talk and optical loss in the 0.7μm pixel. In low-light environments, the GH1 makes use of pixel-merging Tetracell technology that enables higher light sensitivity equivalent to that of a 1.4μm-pixel image sensor.

For video recording at 4K (3,840×2,160) resolutions, the ISOCELL Slim GH1, with a high resolution of 7,968×5,480, is able to take 4K videos with minimum loss in FoV, while most high-resolution image sensors crop or scale down full image resolutions that result in a reduced FoV. Using Tetracell technology, the GH1 is converted down to 3,984×2,740 at 60fps.

GH1 supports a gyro-based EIS, a phase detection AF, and a real-time HDR. The ISOCELL Slim GH1 is expected to be in mass production by the end of this year.

Monday, September 23, 2019

Self-Reset HDR Pixel Paper

A team of researches from Institut für Mikroelektronik Stuttgart, Germany, has chosen quite an unusual place for its paper on HDR pixel - Advances in Radio Science Journal. The paper "Realization and opto-electronic Characterization of linear Self-Reset Pixel Cells for a high dynamic CMOS Image Sensor" by Stefan Hirsch, Markus Strobel, Wolfram Klingler, Jan Dirk Schulze Spüntrup, Zili Yu, and Joachim N. Burghartz "focuses on two topics: One is the realization and opto-electronic characterization of further self-reset pixel cells for an experimental optimization of the functionality with respect to linearity and high signal-to-noise ratio. The second one is the assembly and digital readout of a cluster structure composed of 16 × 16 pixel matrix on a CMOS test chip."

Sunday, September 22, 2019

RED Camera Sensor Conspiracy Theories

Jinni.Tech publishes its investigation of who supplies RED camera sensors:

Thanks to ED for the pointer!

Epticore and Opnous

There is a number of ToF startups in China. Opnous has been been mentioned here a couple of months ago. Apparently, the company has licensed Brookman (Japan) ToF sensor. A recent company presentation tells more about the company technology and differentiation:

Epticore Microelectronics is another startup in China developing ToF technology. The company's presentation shows its products and plans:

Saturday, September 21, 2019

WLO for Next Generation Under-Display Fingerprint Modules

IFNews: GF Securities report on CIS industry shows the next generation under-display fingerprint module that uses WLO to reduce thickness. GF Securities forecasts the ultra-thin screen fingerprint sensors shipments of 94M units in 2020, accounting for 7% of global smartphone shipments, respectively.

Espros pToF Presentation

Espros presentation at MEMS Consulting seminar in China gives interesting info on the company's long range solution for automotive LiDARs:

Friday, September 20, 2019

ToF Sensors in Mobile Devices

TheElec reports that Sony ToF sensors inside LG Innotek modules will be used in Apple 2020 models of iPad and iPhone. Samsung Galaxy S10 5G and Galaxy Note 10+ use Sony ToF sensors too.

LG G8 ThinQ smartphones use ToF sensor from PMD-Infineon combined with ams VCSEL: