History: Hard Battle Against Dust

History: Hard Battle Against Dust

ElectronicsWeekly quotes Sony history site on the yield of early CCDs:

"In January 1980, six years and three months after Iwama first gave orders to begin work on the CCD, the world's first CCD camera was produced. However, the production line's yield of viable CCD chips at this stage was poor and only one in several hundred was usable. It took twelve months to manufacture fifty-two CCD chips necessary to manufacture twenty-six cameras. As a result, each CCD was priced at 317,000 yen. Such a CCD would never realize the 50,000 yen target price that Iwama had specified.

"We can't call this a yield rate. It's more like an occurrence rate of usable chips," complained those overseeing the CCD production. The vicious "enemies" undermining their efforts were superfine dust particles that measured less than several microns and were undetectable by human eye. Clean rooms and dust proof attire reminiscent of space suits were introduced at all development and production sites. The battle against small dust was vigorous; people, machines, and every other thinkable source of dust were identified and dealt with.

After all the measures were applied, the yield rate improved enough to implement full mass production from 1983 at the Kokubu Semiconductor plant in Kagoshima.
"

"Once a quality problem affects an OEM customer, it requires a lot of time and effort to regain that customer's trust. Ochi and his team did have a very difficult time selling the CCD chips in 1983 when they first began OEM sales. They visited various camera manufacturers, but were turned away due to a poor track record. Sony had reduced the volume of semiconductors shipped to its customers to meet a surge in internal demand. Ochi and his team started by apologizing for this. The marketing of CCD chips was literally built on the team's sweat and tears."
Videantis Takeaways from AutoSens 2016

Videantis Takeaways from AutoSens 2016

Videantis posted its review of AutoSens 2016 conference. There are 5 things learned from the conference:
  1. Self-driving cars are hard. After Tesla Autopilot accidents people are more conservative about self-driving prospects. One analyst mentioned 2040 as a possibility for driverless cars to become available, beyond the horizon for most technology companies.
  2. Deep learning is hard. The nets need to be trained with huge amounts of image data, which needs to be properly annotated by hand, with some companies having hundreds of people on staff to perform this task.
  3. Image quality is hard. Image quality will remain a key topic for quite some time.
  4. We need more sensors. OEMs designing 12 cameras into their cars, and this number continuous to go up.
  5. Surround view replacing rear view. These systems are quickly becoming the rear view camera of yesteryear.
First Visual Innovation Award

First Visual Innovation Award

Arizona State University: The new Visual Innovation Award has been announced at IEEE International Conference on Image Processing.

Ren Ng, founder of Lytro light-field camera, has been named the first recipient of the Award. Other finalists related to the image sensing field were Achin Bhowmik, Intel VP, development and deployment lead for Intel RealSense camera technology; Brendan Iribe, co-founder and CEO of Oculus, VR; and Alex Kipman, inventor, Microsoft Kinect.
Nvidia on Key Issues in Automotive Imaging

Nvidia on Key Issues in Automotive Imaging

Image Sensors Auto US publishes an interview with Nvidia imaging architect Joshua Wise. Few points from the interview:

Q: What are the key standards issues need to be addressed as components get more complex and diverse?

Joshua Wise: There are two that are top on my list right now. The first that comes to mind is safety and compliance: an important issue in the automotive environment is the ability to self-diagnose issues. In short, the system must “know when it doesn’t know”. As more components enter the ecosystem, there is more opportunity for data to be damaged in transit — and, similarly, more components result in more health data that must be aggregated and transmitted.

The second on my list is the need for a standard in transmitting data with a high dynamic range. There are as many implementations of sending pixels with greater than 12 bits of data as there are vendors right now — perhaps even more! We’ve been working with vendors to come up with solutions that work with both modern and legacy components, but we see an opportunity to unify and standardize here.

Keynote on Fast Image Sensors

University of Strasbourg and CNRS Prof. Wilfried Uhring presented his keynote "High Speed Image sensors" at SIGNAL 2016 conference on June 27, 2016 in Lisbon, Portugal. Few slides out of 33:


Talking about the IO speed, 25GPixel/s limitation is somewhat obsolete by now. For example, PCIe 4.0 standard defines 32 lanes of 16Gbps each, with the aggregate bandwidth of 512Gbps. Assuming 10b per pixel, one can get 51GPixel/s I/O speed by just buying the PCIe 4.0-compliant IP. And PCIe 4.0 is not the fastest interface these days.

Challenges in Time Correlated Single Photon Counting Imagers

C. Bruschini and E. Charbon (EPFL and Delft TU) presented "(Challenges in) Time Correlated Single Photon Counting Imagers" at SIGNAL 2016 conference held on June 26-30, 2016 in Lisbon, Portugal. Few slides out of 55:

CEVA Presents its XM6 Embedded Vision Platform

CEVA introduces a new DSP-based platform bringing deep learning and AI capabilities to low-power embedded systems. The new IP platform that is centered around a new XM6 imaging and vision DSP:


DJI Mavic Drone Features (Somewhat) Autonomous Flying

SiliconRepublic: DJI Mavic drone uses 5 vision sensors, Movidius Myriad 2 vision processor and 24 processing cores to offer a limited flight autonomy:

Low-Power Event-Driven Image Sensor Architectures

"Low-Power Event-driven Image Sensor Architectures" presentation by Laurent Fesquet, Amani Darwish, and Gilles Sicard of University Grenoble Alpes and CEA-LETI reviews different enent-driven approaches. The presentation was prepared for First International Conference on Advances in Signal, Image and Video Processing held on June 26 - 30, 2016 - Lisbon, Portugal. Few slides from the presentation:

Toshiba Develops Switchable 2D/3D Sensor

Toshiba develops an image sensor that can be switched between 2D and 3D imaging modes. To get a switchable microlens array for 3D light field mode, Toshiba sensor uses a liquid crystal for 8,000 gradient-index (GRIN) lens array. The lenses have 50um pitch and can be switched between 2MP-resolving 2D and a depth-resolving 3D modes by application of +/-4V control voltage:

Sunny Report

Guosen Securities publishes its research report on Sunny Optical - Chinese maker of lens and camera modules. Few quotes:

"Aside from rapid pixel migration (8MP & above over 70% in 1H16), Sunny will benefit from the rapid adoption of dual camera modules in FY16-18. In 2016, we already see Huawei, LG and Xiaomi adopting dual cameras ahead of Apple and Samsung and we expect other competitors very likely to adopt the feature in their upcoming flagship phones to stay competitive. We expect 4% adoption in FY16 rising to 23% through FY18. Beyond that, we predict smartphones will adopt even more cameras to support 3D and augmented reality functions. We forecast HCM revenue to rise at 24% CAGR through FY16-18 while margins should be on a mild uptrend supported by dual camera ramp up."

Snapchat Launches Spectacles with Camera

BBC, WSJ, Business Insider, Snap: Snapchat launches its first hardware product, a camera equipped sunglasses named Spectacles. Simultaneously the company changes its name to Snap to reflect that its activity expands beyond the smartphone application. Now, Snap home page says:

Snap CEO Evan Spiegel wearing Spectacles

Spectacles’ camera uses a 115-degree lens, said to be much closer to the eyes’ natural field of view. On a tap on the button near the hinge, it records up to 10 sec of video from your first-person vantage (30 sec, according to BBC). Each new tap records another clip. Spectacles will be available this fall for $130.

One of the differentiators of the Spectacles camera is the round format of its video output. The rectangular format is said to be the one of the past when photos were made to print on paper. Now, they are used in a different way:



Strange Applications for Image Sensors

Inspired by Ig Nobel prizes this week, here is a roundup of the recent prize-worthy patent applications that incorporate image sensors:

Philips application US20160262521 "A system and a method for guiding a user during a shaving procedure" by Jacek Lukasz Kustra, Guido Theodorus Gerardus Volleberg, Stefan Pfundtner, Karl Catharina Bree, Petrus Johannes Arnoldus Hubertus Kuijten, John Cornelius Petrus Van Der Ven, Harry Broers proposes "the system comprising an imaging module configured to register an image of a part of the body of the user, an image data analyzer configured to determine a local hair-growth direction based on data in the image, a controller configured to generate instructions about a direction in which a hair cutting device is to be moved over the body in dependence on the determined local hair-growth direction, and a feedback system configured to provide the instructions to the user."


Samsung patent application US20160267895 "Electronic device, method for recognizing playing of string instrument in electronic device, and method for providng feedback on playing of string instrument in electronic device" by Dae Young Jeon, Yeon Su Kim, Yeong Min Kim, Jung Min Park, Kyung Seok Oh proposes "The electronic device includes an image sensor configured to sense a motion of a bow to the string instrument, a vibration sensor configured to sense a vibration generated by the string instrument, and a control module configured to determine a fingering position of a user with respect to the string instrument using the motion of the bow and the vibration."


Bosch Und Siemens Hausgerate patent application US20160201981 "Refrigeration appliance with a camera module" by Stephan Kempfle, Gerhard Wetzl solves a problem of rolling shutter distortions when the refrigerator door opens. The distortions might cause difficulties in the food identification. The application proposes to rotate the camera as the door opens, so that it's always pointing to the same spot 125:

Framos Becomes Image Sensor Manufacturer

Framos Becomes Image Sensor Manufacturer

The world's largest image sensor distributor Framos discovers a new opportunity and comes up with its own brand image sensors:

The technological change from CCD to CMOS sensors also offers huge opportunities in medical diagnostics, however, there are no standard sensors for endoscopes. Mostly, customized sensors with a production quantity under 50.000 pieces are not rentable in production due to high on-off-costs.

The imaging experts at FRAMOS are now closing that gap. As a technical consultant and development partner for many years, FRAMOS will launch its own image sensors, especially designed for the needs of the medical industry and endoscopy at the MEDICA 2016. Axel Krepil, Head of the Sensor Department, explains the motivation: “Based on developments in the mobile device market, CMOS technology now offers a clearly superior image quality than CCD in the low-light range relevant for diagnostics. Smaller sensors and thus, endoscope sizes, can be implemented at a very cost-effective price now. As a sensor expert, we can offer such a cost-effective standard product which is qualified for medical use and which offers medical device companies security with respect to development, certification and distribution due to its long-term availability.

The 2MP model FRA0261 with CMOS-BSI technology and with a chip scale packaging (CSP) of 4.1mm x 3.9mm, allows for endoscopes or camera modules with dimensions below 6mm and, at this size, are particularly suited for colonoscopes, mobile and reusable endoscopes used for colonoscopies. A pixel size of 1.4 µm x 1.4 µm, Full-HD resolution and a frame rate of 60 frames per second increase the image quality for the treating physician and improves real-time capabilities. In comparison to CCD technology, the small-size CMOS sensors are available for approximately half the current market price. In addition, FRAMOS offers the sensors as a “die” version with which endoscope manufacturers can install the sensor for a better angle of sight based on optical requirements, in any position, for instance vertical or tilted. With the powerful 2- or 4-lane MIPI interface, only a few cable connections are required between the sensor and the control unit.

The FRA0100 model (in development) with 1MP resolution and a 3mm x 3mm CSP is particularly light and suitable for round small-size endoscopes used in cardiovascular surgeries. The 1000 x 1000 pixel resolution is achieved due to the small pixel size of 1.75µm x 1.75µm.

With both sensors, we cover the need, with respect to reusable sensors which are required in smaller quantities and therefore fall below the radar of the large manufacturers or which would incur costs in individual developments. FRAMOS thus offers a quality adapted to the needs of the medical technology field and a long-term, stable supply at a reasonable price.” says Krepil.

Thanks to TL for the link!
DALSA Introduces TDI CMOS Sensors

DALSA Introduces TDI CMOS Sensors

Teledyne DALSA announces the multi-line color CMOS TDI cameras in its Piranha XL family. Piranha XL cameras feature multi-line color CMOS TDI and employ Teledyne DALSA's CMOS architecture for color imaging using 4 rows per RGB channel for 12 rows in total.

Key features:

  • High sensitivity color TDI for high-speed color imaging with 4x the responsivity of a single-line tri-linear camera through TDI summing of 4 rows per color making it ideal for low-light conditions
  • True RGB colors – outputs native colors without interpolation
  • High speed and high resolution – inspects more in less time with a maximum line rate of 210kHz (70kHz x 3) in full color mode in either 8k or 16k resolution
  • Parallax and sub-pixel spatial correction for ease of integration in challenging environments

Softkinetic Announces World's Smallest ToF Camera Module for Mobile

PRNewswire: SoftKinetic announces the DepthSense 541, said to be the world's smallest 3D ToF camera created for mobile platforms.

"Mobile technology innovation, combined with increased consumer expectations, has created a heightened demand for 3D vision innovation in handheld devices and wearables," said Eric Krzeslo, CMO and co-founder of SoftKinetic. "The rise of location-based augmented reality games requires 3D technology at a smaller footprint, with increased sophistication and capability. SoftKinetic has led innovation in ToF technology and we're excited to continue offer the best 3D depth sensing hardware and software solutions for a wide variety of platforms."

SoftKinetic's DS541 is based on the latest DepthSense ToF architecture and features:

  • 10µ pixel, ¼ inch sensor with micro lenses for highest depth resolution
  • HDR sensor for both outdoor operation and extra close depth capture
  • MIPI interface for lowest latency
  • Innovative lens design for smallest form factor and highest efficiency

The DS541 is currently available for sample, with mass production planned in Q3 2017.


Allied Market Research forecasts the global 3D camera market to reach $10.8 billion by 2021, led by significant embedded adoption in the mobile space, particularly around media and entertainment applications:

3D Camera Market Revenue by Technology, 2014 & 2021

IHS: Cameras Are ~9% of iPhone 7 BOM

IHS publishes its estimation of iPhone 7 BOM, with cameras costing $19.90 or about 8.9% of the total $224.80.

For comparison, iPhone 5s cameras cost was 6-7% of the BOM, iPhone 6 and 6s - about 7.7%, and iPhone SE - about 8.5%.

Image Sensor Americas Agenda

Image Sensor Americas Agenda

Image Sensor Americas conference to be held on Oct. 25-26 in San Francisco, CA, publishes its quite an interesting agenda with short abstracts:
  • Keynote: The Prospects for Building Image Sensors in America
    Paul Gallagher | Senior Director of Technolog of LG Display
  • Camera Selection for Low Light Imaging: Choosing the Right Camera Based on an Informed Analysis of Signal Levels and Noise.
    James Butler | Manager of OEM Camera Products of Hamamatsu Photonics
  • Image Sensors for Security, Wearable and Consumer Electronic Products: The Race to Low Light Performance
    Patrice Roulet | Director of Engineering and Co-Founder of Technology of Immervision
  • Fueling an IoT-Driven Future: The QuantumFilm Imaging Platform
    Remi Lacombe | VP Sales and Marketing of Invisage
  • Flexible Image Sensors Printed on Plastic
    Carlo Guareschi | Vice-President of Business Development of ISORG
  • Bio-inspired Event-based Sensor Technology for Computer Vision
    Luca Verre | CEO of Chronocam
  • Imaging beyond the Bandgap: A 200 GHz Silicon CMOS Imager
    Kenneth Fourspring | Senior Image Scientist of Harris Corporation Geospatial Systems
  • Processors for Embedded Computer Vision: Options and Trends
    Jeff Bier | Founder and President of Embedded Vision Alliance and BDTI
  • CMOS Image Sensors Open Up New Options in Medical Imaging
    Axel Krepil | Sensor + Division of Framos
  • Miniature Form Factor Camera Modules for Medical Endoscopy
    Martin Waeny | CEO of AWAIBA
  • Market Analysis
    Rajender Thusu | Indsutry Principal, Sensors Practice of Frost & Sullivan
  • Optical Solution for NIR Crosstalk in RGBir Hybrid Sensors
    Rich Hicks | Senior Camera and Imaging Technologies of Intel
  • ALCHERA: a hyperspectral architecture that enables ubiquitous imaging spectroscopy
    Alex Hegyi | Research Staff of PARC, a Xerox Company
  • Using Advanced Imaging to Optimise Supply Chain Processes: The Beef Example
    Alexandra Booth | Research Scientist of ImpactVision
  • Image Sensor Needs: A Customer’s Perspective
    Scott Campbell | V.P. of Research of GoPro
  • The Eyes Have It
    Peter Milford | CTO of EyeFluence
  • IEEE CPIQ Standard for Measuring and Benchmarking Image Quality in Consumer Mobile Devices
    Jonathan Phillips | Image Scientist of Google
  • Mobile Imaging
    Sean Kelly | VP of Imaging of Motorola
  • Deep Learning, Driving the next Image Processing Revolution
    Yair Siegel | Director of Product Marketing, Imaging & Vision of CEVA DSP
  • Using Computational Power to Overcome the Hardware Constraints in Optical Designs
    Eugene Panich | CEO, Co-Founder of Almalence
  • Image Sensors in Automotive: Plotting the Evolution of Image Sensors in a Vehicle
    Pallavi Madakasira | Analyst of Lux Research Intelligence

Omnivision Launches 120dB 1.7MP Automotive Sensor

PRNewswire: OmniVision announces the OV10650, a wide-format imager with 2:1 aspect ratio. Built on 4.2um OmniBSI split pixel, the OV10650 delivers 1820 x 940 resolution at up to 60fps and 120dB of DR.

"Offering a wide field of view and high dynamic range, we view the OV10650 as ideally suited for next-generation rear video mirrors utilizing wide high-resolution displays," said Thilo Rausch, senior automotive product manager at OmniVision. "As a result, the driver can make better, more informed decisions, especially in challenging driving conditions due to increased situational and spatial awareness around the vehicle."

OmniVision is also launching the OV10652, same as the OV10650, but utilizes an RCCC color filter pattern for automotive camera integrators for ADAS. Both OV10650 and OV10652 are specified to meet the standards required by the 2018 New Car Assessment Program (NCAP). Systems based on these sensors can enable features such as pedestrian detection, lane departure warning, head lamp control, forward collision warning, and traffic sign detection.

Omnivision Launches VGA GS Sensor

PRNewswire: OmniVision announces the latest addition to its portfolio of automotive image sensors, the 3um global shutter pixel-based OV7261 is aimed to gesture control and driver monitoring applications such as driver drowsiness and distraction detection.

"We project that the driver state monitoring system market will grow significantly through 2020, which underscores the automotive industry's drive to implement advanced camera applications not only on high-end luxury vehicles, but also on mainstream automobiles across multiple markets," said Jeffrey Morin, senior automotive marketing manager for North America at OmniVision. "The OV7261 aims to drive down the cost, power, and space requirements of implementing vision-based driver monitoring systems without sacrificing image quality or product reliability."

ON Semi Unveils 8MP Sub-Electron Noise EMCCD

BusinessWire: The new 8MP 4/3-inch KAE-08151 IT-EMCCD delivers the sub-electron noise floor for applications like high resolution microscopy and scientific imaging operating in lighting regimes from sub-lux to bright-light imaging. In addition, a new packaging option is available which incorporates a thermoelectric cooler (TEC) directly into the package design. This integrated cooler simplifies development of a cooled camera that optimizes the performance available from these devices.

The compelling combination of performance and flexibility provided by IT-EMCCD technology is providing significant benefits for low-light imaging across key industrial imaging applications,” stated Herb Erhardt, VP and General Manager, Industrial Solutions Division, Image Sensor Group at ON Semiconductor. “By expanding our product portfolio with a new resolution node and options that simplify the integration process for camera manufacturers, end customers will be presented with even more ways to leverage the advantages of this unique technology.

The KAE-08151 is sampling today in Monochrome and Bayer Color configurations in a CPGA-155 package, with samples incorporating an integrated TEC available in the first quarter of 2017.