Sunday, May 31, 2015

FrinGOe Unveils Consumer FT Spectrometer

FrinGOe announces a cheap Fourier Transform spectrometer of consumer applications. The optical part of the spectrometer is a passive add-on for iPhone or Raspberry Pi camera. The company's whitepaper explains how it works:

Regular FTIR spectrometer
Mach-Zehnder FTIR spectrometer with moving mirrors
Passive array of Mach Zehnder interferometers with
path difference fixed at different values.

"FrinGOe uses innovative crystal technology to create a low cost 2D array of Mach Zehnder Interferometers and patented optical configuration to map the array onto the camera sensor. This allows an all-passive and compact optical add-on (as small as 4mm by 4mm by 2mm) that can be mounted onto any imaging platforms such as a mobile phone camera, Raspberry Pi camera or other OEM devices." One of the prototypes of the FT spectrometer has a spectral resolution of ~1 nm @ 400nm wavelength.

FrinGOe on the back of iPhone 5s is switchable between spectroscopic
and imaging modes

In June, the company plans to launch FrinGOe at the Laser World of Photonics (Munich) exhibition and at the same time start an Indiegogo campaign to kick-start the mass production efforts. FrinGOe's Youtube video shows the spectrometer operation:

Friday, May 29, 2015

Apple Acquires Metaio

Techcrunch reports that Apple has apparently acquired a German AR company Metaio. The company has a number of AR projects, one of them being Thermal Touch covered in this blog a year ago. Another interesting project out of many was an AR application for camera-equipped smart watch (Youtube link):

Target and Carrefour Test VLC

Electronics360: Carrefour SA, the largest retailer in Europe, is experimenting with visible light communications, becoming the first major retailer to install Philips’ connected lighting system with LED-based indoor positioning at its newly refurbished supermarket in Lille, France. The Carrefour news comes about a month after it was reported that U.S. retailer Target will use indoor positioning, widely believed to be VLC-based, to power its in-store app to help customers navigate its stores.

The Carrefour supermarket system has 800 LED fixtures that use Philips VLC technology to transmit location. Philips software and cloud-based location database are integrated into Carrefour’s mobile app.

Update: Philips posted infographics of how it works:

Philips also posted a Youtube video showing its Carrefour trial:

We are leading the way with connected lighting for retail with Carrefour,” comments Gerben van der Lugt, Head of LED-based indoor positioning at Philips Lighting. “Our connected lighting system has the potential to transform shopping into a more interactive and personalized experience. At the same time it will enable retailers to differentiate themselves, enhance customer loyalty and provide new services to shoppers.

Omnivision Reports Quarterly and Yearly Results

PRNewswire: Omnivision reports quarterly results for its fiscal quarter ended on April 30, 2015.

Revenues for the fiscal year ended April 30, 2015 were $1.4 billion, as compared to $1.5 billion in fiscal 2014. GAAP net income for fiscal 2015 was $93.4 million, as compared to $95.0 million a year ago.

Revenues for the last quarter were $285.9 million, as compared to $292.3 million in the previous quarter, and $331.0 million a year ago. GAAP net income in the last quarter was $6.0 million, as compared to net income of $14.0 million in the previous quarter, and $15.1 million a year ago. GAAP gross margin for the last quarter was 22.4%, as compared to 22.1% for the previous quarter and 20.1% a year ago. The Company ended the period with cash, cash equivalents and short-term investments totaling $524.2 million, an increase of $11.4 million from the previous quarter.

Based on current trends, the Company expects revenues for the next quarter will be in the range of $310 million to $340 million.

SeekingAlpha publishes a trascript of Omnivision's earnings call. There was no Q&A session for some reason. Shaw Hong, Chairman and CEO, says on the acquisition news: "by partnering with the consortium and operating as a private company, we will have more flexibility to innovate and develop new solutions for our customers, which will benefit our employees, customers and business partners. We expect the transaction will close within the next nine to 12 months.

As of today, we have not obtained all the necessary approval, nor have we met the closing conditions. As such, there's no assurance that the proposed acquisition by the consortium will close.

Thursday, May 28, 2015

Google Presents VR Shooting Setup with 16 Cameras

Google presents 16-GoPro-cameras JUMP setup for shooting VR video at its I/O conference:

Cnet video shows how it works:

Pelican Imaging Expands into Array of Different Cameras

Up to now, Pelican Imaging camera components in the array had the same FOV and viewing direction. Its new patent application US20150146029 "Array Camera Configurations Incorporating Multiple Constituent Array Cameras" by Kartik Venkataraman and Jacques Duparré, presents integration of "a plurality of constituent array cameras, where each constituent array camera has a distinct field of view and/or a distinct viewing direction." The main advantage of that is implementation of zoom feature:

Now it becomes fairly similar to Corephotonics and Linx Imaging proposals.

Ximea on Hyperspectral Imaging

German machine vision camera vendor Ximea CEO Max Larin discusses hyperspectral imagers approaches and applications in this Youtube video:

Wednesday, May 27, 2015

Sony 2015 IR Day

Sony has held a 2015 Investor Relations Day today. Device Segment presentation by Tomoyuki Suzuki, Executive Deputy President and Corporate Executive Officer responsible for the Device business, has a lot of info on Sony image sensor business:

Samsung Presents 1um RWB Pixels

PhoneArena quotes Chinese-language posting pictures from Samsung image sensor seminar. Samsung presents its 1um RWB color patterned sensors, said to provide much better low light sensitivity:

Tuesday, May 26, 2015

Canon vs IV Patent Arguments

Delaware Intellectual Property Litigation publishes a detailed document on trial where Intellectual Ventures accuses Canon in patent infringement, including two image sensor patents: US6,221,686 "Method of making a semiconductor image sensor" and US6,023,081 "Semiconductor image sensor." Canon has lost the battle, in spite of Eric Fossum and Albert Theuwissen trying to defend it, see pp. 11-19 in the document.

Hamamatsu Image Sensor Sales

Hamamatsu reports its quarterly results, including image sensor sales:

Monday, May 25, 2015

Assorted Market Data

China Galaxy International comes with a wide collection of market data, primarily about Apple camera module supplier Cowell and the camera module industry in general:

Audi Parners with Gestigon to Develop Gesture Control in Cars

gestigon announces that they provide the gesture control software used in Audi’s booth at the CES ASIA in Shanghai. The first result of this cooperation is to, initially, provide an unique user experience for the rear seat passengers of Audi’s luxury car models enabling the user interaction with the multimedia device in front of them.

Moritz von Grotthuss, CEO of gestigon, says, “we are excited to have won AUDI AG as a partner to develop gesture control solutions for automotive. The application for the rear passengers is a first, appealing step into a multitude of use cases based on ‘active control’ to enable new, innovative and easy-to-use functionalities. With Audi’s unique positioning as one of the leading global premium manufacturers we have found a partner that shares our vision and excitement in regard of skeleton tracking- and gesture control for automotive in their broadest sense.

Michael Schlittenbauer, Developer at AUDI AG confirms, „gestigon was able to provide a customized software solution for our targeted use case in the rear of the cabin. We are working on a number of concepts to integrate gesture control into our models and are looking forward to continue to explore the technology as well as use cases to enable a unique driving experience to our customers.

Saturday, May 23, 2015

Sony and Apple Think Alike

Sony patent application US20150129943 "Solid-state imaging device, method of manufacturing a solid-state imaging device, and electronic apparatus" by Nanako Kato and Toshifumi Wakano, presents a stacked global shutter pixel where the storage diode is shielded from the backside illumination:

Sony stacked GS pixel

Sony's idea looks quite similar to the one presented by Apple in US20150035028 application:

Apple stacked GS pixel

Friday, May 22, 2015

IC Insights: Sharp CIS Sales Surge

IC Insights posts its updated list of top 20 semiconductor companies. This time, Sharp has entered the list as #19. IC Insight writes: "Although Sharp as a whole is having a difficult time, its semiconductor group, which represents only about 14% of the company’s corporate sales, posted a whopping 62% growth rate (an 88% increase in yen), the best 1Q15 sales increase of any top-20 semiconductor supplier. This sales surge was almost entirely due to the company’s success in the CMOS image sensor market."

Update: Sharp's medium term management plan dated by May 14, 2015, reveals quite ambitious plans over the next few years:

The company notes: "Sharp has the leading share in the industry for smartphone camera devices, but since specific customers weight a large proportion of our business, we will aggressively work to expand our sales to new customers, and target 3 times the net sales of Fiscal 2014 by Fiscal 2017."

Invisage Videos Deleted

Invisage has removed all its nice videos from Youtube. There is only one left in the Invisage Tube channel:

Thursday, May 21, 2015

Toshiba Image Sensor Circus Continues

Toshiba updates its popular explanations of image sensor work it started in 2013. The new CMOS Image Circus series compares image sensor with human eye, talk about image sensor features and more:

PMD Announces ToF Camera for AR and VR Devices

PMD announces pico flexx, a version of pico ToF camera optimized for augmented and virtual reality devices, featuring higher speed, longer distance and lower power:

PMD also publishes a Youtube demo of the new camera:

Wednesday, May 20, 2015

IHS on Omnivision Acquisition

IHS publishes its analysts opinions on implications of Omnivision acquisition by Chinese investment companies. Few quotes:

"One of the greatest concerns regarding China-backed Hua Capital Management Co. Ltd.’s proposed acquisition of OmniVision Technologies Inc. is that it would give OmniVision an unfair advantage over other image-sensor vendors, with Chinese firms favoring Omnivision products over others. But, according to IHS, this scenario is unlikely to play out.

...the vast majority of OmniVision’s current business is already in China. O’Rourke says about 75% of OmniVision’s 2014 revenues came from China-based companies.

“In this situation, there’s not a great deal of room left for OmniVision’s growth in China," O’Rourke says. "Thus, the question of Chinese companies or the Chinese government favoring it may be irrelevant.”

...the funding groups bought OmniVision mainly for a “capital operation” and will have the company listed in the China stock market to gain a sizable return.

Wang says he believes OmniVision will invest heavily in the near term—including spending more on R&D to develop high-end products—in order to help the company catch up with image sensor heavyweights Sony and Samsung.

Sony Presents New Generation Global Shutter Sensors with 3.45um Pixel

Sony announces 5.07MP 2/3-inch IMX250LLR/LQR and 3.2MP 1/1.8-inch IMX252LLR/LQR sensors featuring a new generation global shutter pixel. Sony says that the new 3.45um pixels are industry's smallest global shutter pixels for industrial applications. The pixels feature improved sensitivity, 1.1 times over the company's previous generation of 5.86 µm pixels. In addition, EXview HAD CCD technology is said to increases sensitivity in NIR (see the bigger 800 x 670 pictures on Sony site):

The sensors are fairly fast, slightly faster than the previous generation IMX174:

Sony also publishes a Youtube video showing a global shutter advantage:

Pelican Gets New COO

PRNewswire: Pelican Imaging announces that Mark Fulks has joined the company as COO to lead engineering and product development for Pelican's depth-sensing array solutions. Prior to joining Pelican, Fulks led Nokia's product and engineering teams that created Nokia's augmented reality product, Nokia City Lens, and led the core development teams at Sun Microsystems for Java ME (J2ME). He began his career as an embedded systems software engineer, after earning his Bachelor's degree in Computer Science with a minor in Physics from Sonoma State University. Fulks holds several patents related to augmented reality interfaces and displays.

"Pelican Imaging is uniquely positioned to achieve success across multiple growth markets. I'm very excited to join as COO, leading the engineering and product development teams to bring this disruptive 3D imaging technology to consumers," said Fulks.

Tuesday, May 19, 2015

e2v Reports Yearly Results

e2v reports preliminary results for its fiscal year ended on March 31, 2015. For the imaging business, the report states:

Steve Blair, e2v CEO, comments on the last year's Anafocus acquisition "We are pleased with the performance of AnaFocus since its acquisition in September and it has already been successfully integrated. Overall, we have delivered a good performance, including a strong finish to the year, in what remained challenging markets."

Monday, May 18, 2015

Omnivision News: MIPI over Fiber, 3DInCites Award Nomination

PRNewswire: OmniVision and Silicon Line GmbH announced a collaboration to bring high-speed image and video data transfer technology to applications that are noise-sensitive and that require longer, thinner and lighter cables. This collaboration will focus on 1080p30 video transfer over 2-lane 1Gbps MIPI for OmniVision's OV5693 sensor, as well as other MIPI D-PHY 1.1, 1.2, 2.0 and MIPI C-PHY-based imaging solutions.

"We believe that Silicon Line's approach to high-speed data transfer, and the inherent benefits that its optical link technology brings to our mutual target markets, can help overcome many technical challenges faced by product engineers today," said Tehzeeb Gunja, senior marketing and business development manager at OmniVision. "The ability to affordably and power-efficiently transfer large amounts of uncompressed video data over long distances carries a tremendous potential for positive impact on a number of application areas."

"As OmniVision extends its portfolio with higher resolution and higher frame rate image sensors, the required data rate for transporting the image increases dramatically," said Ian Jackson, senior manager sales and marketing at Silicon Line. "A single optical fiber can transport up to 25 Gbps of data. This has obvious benefits in terms of size and weight reductions, enabling smaller and lighter industrial designs for space-constrained applications like tablets and notebooks. Another benefit of optical fiber is zero susceptibility to electromagnetic interference, which enables a noise free picture. Lastly, optical technology can transport high-speed image data over many meters whereas a traditional electrical connection is limited to 10 or 20 cm, making optical technology well-suited for applications such as automotive, medical and security imaging systems." (Actually, a single optical fiber transmits 100Gbps since 2012 in commercial products, and, this year, 200Gbps would be probably commercialized - ISW.)

3DInCites nominates OmniVision’s 23.8MP OV23850 PureCel stacked sensor for 3D Excellence Award. OV23850 "gives 3D manufacturing visibility across numerous application areas, including automotive, medical, mobile devices, and wearables."

3D News: Orbbec, Chiaro, OptAssyst

Shenzhen, China-based Orbbec starts to accept pre-orders for its structured light 3D cameras: a longer range Astra, and a shorter range Astra S. Both Astra cameras are supposed to be compatible with discontinued after acquisition by Apple Primesense cameras and its NUI middleware:

Orbbec comes up with a series of comparison videos against Primesense, Kinect-1 and Kinect-2 cameras.

Thanks to HJ for the pointer.

Boulder, CO-based Chiaro Technologies presents its 3D sensing technology based on patented Symbolic Light. Chiaro's camera consists of a sensor head and a decoding API for either a PC host or an embedded system. The efficient 3D decoding improves frame-rate while leaving a room for application. Cloudburst is designed with machine vision in mind.

Chiaro camera is said to excel in building 3D maps of reflective objects, such as metallic ones:

Two coins and a machined part on a black-and-white checker pattern. Under low lights in a 12-bit image (a), specular reflections saturate the camera while the checker pattern remains dim. Typical 3D capture (b) misses regions of the 3D point cloud. Chiaro’s XDI-3D system with HDR (c) can capture 3D images of this scene, capturing everything from the black squares to the shiny surfaces of coins, all under regular ambient lighting conditions.

Meanwhile, German Optassyst drives an ISO initiative "Fair Datasheet" to define a standard way to write datasheets for 3D depth cameras, so that customers can easily compare their performance and merits. The proposed ISO standard definitions mainly target machine vision and industrial 3D cameras.

Thanks to TL for the link!

Sunday, May 17, 2015

Fujitsu Gaze Tracking Technology

Fujitsu publishes few Youtube videos on its newly developed gaze tracking technology. The first one shows a short-range "compact tracker:"

The second video shows a solution for a long-range gaze tracking:

The third video talks about possible applications and markets for the gaze tracking:

Friday, May 15, 2015

Fujitsu Iris-Scanning Smartphones Go On Sale

The Guardian: Fujitsu Arrows NX F-04G smartphone sold by Japan’s NTT Docomo, uses an infrared LED and a special camera to identify a user by snapping a picture of the iris of his eye. The camera was first presented and awarded at MWC 2015. Most implementations of iris-recognition systems are said to work at a distances of 10cm, while Fujitsu’s system can operate at a standard smartphone operating distance. FujitsuMobile shows how it works:

WSJ reports that ZTE too presented eye recognition feature in its new Grand S3 phone at MWC 2015, although it uses a different approach licensed from EyeVerify. It works by imaging the unique vein patterns in the whites of the eye using the Grand S3’s front facing camera:

GizmoChina: ViewSonic too announced its plans on iris recognition in smartphone, but it's not clear if it has reached the market yet.

Meanwhile, Forbes discusses the hacks to fool the iris recognition cameras:

We have managed to fool a commercial system with a print out down to an iris diameter of 75 pixels,” says security researcher Jan “Starbug” Krissler. “I did tests with different people and can say that an iris image with a diameter down to 75 pixel worked on our tests.” The print out had to have a resolution of 1200 dpi too, though it’s easy to find printers able to hit that specification today, and ideally at least 75 per cent of the iris was visible.

Unlike the fingerprint attack, where it was necessary to create a proper clone, all that he needed in his iris recognition hacks was the print out, the researcher claims. “It’s nothing more. I punched a hole in the middle, but only for orientation. It’s not needed,” Starbug added.

Thursday, May 14, 2015

Himax Image Sensor Business Update

Himax quarterly earnings report updates on the company's image sensor business:

"Himax’s CMOS image sensors also experienced a down quarter since 4G smartphone adoption in China remained weak. The lack of smartphone replacement demand hurt the shipments of Himax’s high end product offerings. The Company is optimistic of the possibility of securing significant design-wins in the second half of the year with its 8-megapixel and 13-megapixel sensors."

From the company's May 2015 presentation:

CISTERN and EXIST Programs Support Image Sensor R&D in Europe

Adimec announces its participation in two European projects for advancing CMOS image sensor technology: CISTERN and EXIST. Both programs begin in April/May 2015 and will be completed in 2018.

The main objective of CISTERN, which is an acronym for CMOS Image Sensor Technologies’ Readiness for Next generation of applications. The other partners in this project are Grass Valley (project leader), Thales Angenieux, SoftKinetic, Delft University of Technology, University of Burgundy. CISTERN is a part of CATRENE Program and is aimed to:
  • To develop CMOS imager sensors with improved performance on spatial resolution, temporal resolution, higher bit depths, lower noise, wider color gamut, higher quantum efficiency (not necessarily all combined for each sensor) and additional, smart functionality in the pixels. The consortium will develop imager sensors in the categories UHDTV, Time-of-Flight and hyper- and multi-spectral, for the application areas broadcast, entertainment, and security.
  • To develop real time image processing techniques needed to improve the quality of the digital output signal of the sensor demonstrators.
  • To develop and demonstrate the capability to produce multispectral imagers by hybridization of multispectral filter arrays on top of CMOS sensor.
  • To demonstrate the improved performance of the CMOS imagers combined with related processing in a number of demonstrators.
  • To develop Ultra High resolution, widely opened, sensor adapted zoom lenses (2/3’’ format, 4K resolution) for broadcast and security applications/markets.
  • To optimize lens design in term of performance/cost/weight/camera integration for integral imaging chains.
  • To develop and demonstrate an integrated Camera Lens Assembly for security applications that offers UHDTV performance with Size and Weight of an HDTV solution.
  • To start in-house CMOS image sensor development in Grass Valley.
The EXIST (EXtended Imaging Sensor Technologies) consortium will investigate and develop innovative new technologies for image sensors. The EXIST team consists of R&D institutes advanced in image sensor technologies, image sensor designs and video processing (IMEC, Fraunhofer IMS, Le2i, TNO, iMinds and TU Delft) and fabless design houses, semiconductor manufacturer (Sofradir) and system integrators (CMOSIS, Grass Valley, Softkinetic, Thales Angenieux, Adimec, Silios, Sofradir, Focal, and Quest). The objectives of this program are investigate and develop innovative new technologies for image sensors needed in the next plus one (N+2) generation of several application domains. The image sensor research will focus on enhancing and extending the capabilities of current CMOS imaging devices:
  • New design (architectures) and process technology (e.g. 3D stacking) for better pixels (lower noise, higher dynamic range, higher quantum efficiency, new functionality in the pixel) and more pixels at higher speed (higher spatial and temporal resolutions, higher bit depth), time-of-flight pixels, local (on-chip) processing
  • Extended sensitivity and functionality of the pixels: extension into infrared, filters for hyperspectral and multispectral imaging, better colour filters for a wider colour gamut, and FabryPérot Interference cells
  • Increasing the optical, analog and data imaging pipelines to enable high frame rates, better memory management, etc.
After developing the related technologies, six prototype image sensors will be designed:
  • two high resolution, high frame rate image sensors with 32Mpixels or more;
  • two Time of Flight image sensors;
  • one 3D stacked image sensor.
  • one integrated high performance hyperspectral image sensor
Next to these image sensor designs, the team will make a number of filter designs for integration on top of above mentioned and commercially available image sensors:
  • Multispectral filters for MWIR
  • Hyperspectral filters for NIR-passband and NIR-cutoff filters;
  • HSI filters and BSI imager for VISNIR range;

Panasonic Image Sensor Based Datacom

DigInfo TV publishes a video on Panasonic image sensor optical commonication technology:

Wednesday, May 13, 2015

CogniVue Announces Next Generation Cognition Processor

PRNewswire: CogniVue, which has licensed its APEX Image Cognition Processor (ICP) technology to Freescale, announced that its new APEX Processing Core - Opus - is at the center of the intense battle for ADAS market share as challengers to Mobileye, the current market share leader, line up to capture business for next-generation ADAS systems. Opus is said to have 10x performance increase over the current G2-APEX.

"CogniVue has been anticipating this ADAS market showdown and we are ready to equip competitors to Mobileye with the performance and cost advantages they need to capture new ADAS opportunities," said Simon Morris, CEO of CogniVue. "We have the innovation, technology and engagement with dominant players in the automotive market and key SoC suppliers to be a key enabler of next-generation ADAS and autonomous driving."

Tuesday, May 12, 2015

ST Explores Its Options with Imaging Business

ST has held a 2015 Analyst Day today. The company says it's exploring options to cut losses at its Digital Product Group (DPG), a division that makes image sensors among other products. "This problem needs to be fixed and we will fix it," says CEO Carlo Bozotti (SeekingAlpha).

Meanwhile, DPG presentation says that ST imaging business is focusing on specialized image sensors and proximity sensors, while winding down commodity imaging products:

In spite of the option explorations, the manufacturing presentation talks about imaging and the 3D roadmap: