Hamamatsu publishes two more videos - "LiDAR, Radar, and Cameras: Measuring distance with light in the automotive industry" and "SiPM: Operation, performance, and possible applications," both by Slawomir Piatek, senior lecturer of physics at New Jersey Institute of Technology.
Tuesday, February 13, 2018
Monday, February 12, 2018
Hamamtsu SiPM Theory and Comparisons with PMT
Hamamtsu publishes two hour-long educational webcasts "Silicon photomultipliers: theory & practice" and "Low light detection: PMT vs. SiPM."
Jaroslav Hynecek Elevated to IEEE Fellow
Jaroslav Hynecek has been elevated to IEEE fellow for contributions to solid-state image sensors.
Thanks to NT for the info!
Thanks to NT for the info!
Facial Recognition Glasses for Chinese Police
The Verge reports that China police is testing facial recognition glasses at train stations in the “emerging megacity” of Zhengzhou, where they’ll be used to scan travelers during the upcoming Lunar New Year migration.
The glasses are developed by Beijing-based LLVision Technology Co. The company says they’re able to recognize individuals from a pre-loaded database of 10,000 suspects in just 100ms, but cautions that accuracy levels in real-life usage may vary due “environmental noise.”
The glasses are developed by Beijing-based LLVision Technology Co. The company says they’re able to recognize individuals from a pre-loaded database of 10,000 suspects in just 100ms, but cautions that accuracy levels in real-life usage may vary due “environmental noise.”
Sunday, February 11, 2018
SPAD Sensor for Entangled Photon Imaging
SPIE publishes FBK paper and presentation video of "SUPERTWIN: towards 100kpixel CMOS quantum image sensors for quantum optics applications" by Leonardo Gasparini; Bänz Bessire; Manuel Unternährer; André Stefanov; Dmitri Boiko; Matteo Perenzoni; David Stoppa.
"Quantum imaging uses entangled photons to overcome the limits of a classical-light apparatus in terms of image quality, beating the standard shot-noise limit, and exceeding the Abbe diffraction limit for resolution. In today experiments, the spatial properties of entangled photons are recorded by means of complex and slow setups that include either the motorized scanning of single-pixel single-photon detectors, such as Photo-Multiplier Tubes (PMT) or Silicon Photo- Multipliers (SiPM), or the use of low frame rate intensified CCD cameras. CMOS arrays of Single Photon Avalanche Diodes (SPAD) represent a relatively recent technology that may lead to simpler setups and faster acquisition. They are spatially- and time-resolved single-photon detectors, i.e. they can provide the position within the array and the time of arrival of every detected photon with less than 100 ps resolution. SUPERTWIN is a European H2020 project aiming at developing the technological building blocks (emitter, detector and system) for a new, all solid-state quantum microscope system exploiting entangled photons to overcome the Rayleigh limit, targeting a resolution of 40nm. This work provides the measurement results of the 2nd order cross-correlation function relative to a flux of entangled photon pairs acquired with a fully digital 8×16 pixel SPAD array in CMOS technology. The limitations for application in quantum optics of the employed architecture and of other solutions in the literature will be analyzed, with emphasis on crosstalk. Then, the specifications for a dedicated detector will be given, paving the way for future implementations of 100kpixel Quantum Image Sensors."
The trend in CMOS SPAD-array design goes towards:
(i) the miniaturization of the pixel (below 10µm) to increase the output image resolution;
(ii) SPAD optimization to improve the photon detection efficiency (PDE) while reducing DCR, after-pulsing and crosstalk;
(iii) 3D stacking of chips, with a top tier optimized for sensing that includes the array of SPADs and a bottom tier optimized for processing (i.e., counting, timestamping and buffering);
(iv) smart mechanisms for timestamping photons, such TDC sharing and time-gated counting in the analog domain;
(v) the on-chip implementation of pre-processing stages, such as timestamp histogramming, to reduce the sensor output data size and increase the frame rate, thus enabling synchronization with fast sources of photons (from 100kHz up to tens of MHz).
"Quantum imaging uses entangled photons to overcome the limits of a classical-light apparatus in terms of image quality, beating the standard shot-noise limit, and exceeding the Abbe diffraction limit for resolution. In today experiments, the spatial properties of entangled photons are recorded by means of complex and slow setups that include either the motorized scanning of single-pixel single-photon detectors, such as Photo-Multiplier Tubes (PMT) or Silicon Photo- Multipliers (SiPM), or the use of low frame rate intensified CCD cameras. CMOS arrays of Single Photon Avalanche Diodes (SPAD) represent a relatively recent technology that may lead to simpler setups and faster acquisition. They are spatially- and time-resolved single-photon detectors, i.e. they can provide the position within the array and the time of arrival of every detected photon with less than 100 ps resolution. SUPERTWIN is a European H2020 project aiming at developing the technological building blocks (emitter, detector and system) for a new, all solid-state quantum microscope system exploiting entangled photons to overcome the Rayleigh limit, targeting a resolution of 40nm. This work provides the measurement results of the 2nd order cross-correlation function relative to a flux of entangled photon pairs acquired with a fully digital 8×16 pixel SPAD array in CMOS technology. The limitations for application in quantum optics of the employed architecture and of other solutions in the literature will be analyzed, with emphasis on crosstalk. Then, the specifications for a dedicated detector will be given, paving the way for future implementations of 100kpixel Quantum Image Sensors."
The trend in CMOS SPAD-array design goes towards:
(i) the miniaturization of the pixel (below 10µm) to increase the output image resolution;
(ii) SPAD optimization to improve the photon detection efficiency (PDE) while reducing DCR, after-pulsing and crosstalk;
(iii) 3D stacking of chips, with a top tier optimized for sensing that includes the array of SPADs and a bottom tier optimized for processing (i.e., counting, timestamping and buffering);
(iv) smart mechanisms for timestamping photons, such TDC sharing and time-gated counting in the analog domain;
(v) the on-chip implementation of pre-processing stages, such as timestamp histogramming, to reduce the sensor output data size and increase the frame rate, thus enabling synchronization with fast sources of photons (from 100kHz up to tens of MHz).
Friday, February 09, 2018
Elphel Quad 3D Camera Adds CNN and Tile Processor
Open-source Elphel camera project that uses quad camera for 3D vision, now adds Convolutional Neural Network with intention to reach the depth range of few hundreds to thousands meters with cameras spaced apart by just 150mm:
"We plan to fuse the methods of high resolution images calibration and processing, already emulated functionality of the Tile Processor (TP), RTL code developed for its implementation and the Convolutional Neural Network (CNN). Compared to the CNN alone this approach promises over a hundred times reduction in the number of input features without sacrificing universality of the end-to-end processing. The TP part of the system is responsible for the high resolution aspects of the image acquisition (such as optical aberrations correction and image rectification), preserves deep sub-pixel super-resolution using efficient implementation of the 2-D linear transforms. Tile processor is free of any training, only a few hyperparameters define its operation, all the application-specific processing and “decision making” is delegated to the CNN."
"We plan to fuse the methods of high resolution images calibration and processing, already emulated functionality of the Tile Processor (TP), RTL code developed for its implementation and the Convolutional Neural Network (CNN). Compared to the CNN alone this approach promises over a hundred times reduction in the number of input features without sacrificing universality of the end-to-end processing. The TP part of the system is responsible for the high resolution aspects of the image acquisition (such as optical aberrations correction and image rectification), preserves deep sub-pixel super-resolution using efficient implementation of the 2-D linear transforms. Tile processor is free of any training, only a few hyperparameters define its operation, all the application-specific processing and “decision making” is delegated to the CNN."
SMIC CIS Sales Grow 70% YoY
SeekingAlpha publishes SMIC earnings call transcript with update on its CIS business:
"We have already pinpointed a number of key platforms to address and today I'll highlight two of them. Our NOR flash platform and CMOS image sensor platform. These two have revenue to SMIC grow almost 70% in last year, compared with the year before. We continue to build on our platform strategy and seek to expand our customers' business. We are working hard to implement this market adjustment strategy within the company."
"We have already pinpointed a number of key platforms to address and today I'll highlight two of them. Our NOR flash platform and CMOS image sensor platform. These two have revenue to SMIC grow almost 70% in last year, compared with the year before. We continue to build on our platform strategy and seek to expand our customers' business. We are working hard to implement this market adjustment strategy within the company."
More Details from Sony IEDM 2017 Presentation
Fuse publishes few more slides from Sony presentation on 3-layer chip stacking flow at IEDM 2017.
"The final product is an impressive 19.3M pixels of 1.22 x 1.22 μm each and a 1 Gbit DRAM. Sony used TSVs that have a minimum diameter of 2.5 μm and a pitch of 6.3 μm with a line of 2 μm and space of 0.64 μm. In total they have over 35,000 TSVs – about 15,000 connecting the pixel substrate and the DRAM substrate and about 20,000 more connecting the DRAM substrate to the logic substrate.
The chip achieved 120 fps for all 19.3M pixels and can produce 960 fps FHD (1,920 x 1,080) super slow motion video."
"The final product is an impressive 19.3M pixels of 1.22 x 1.22 μm each and a 1 Gbit DRAM. Sony used TSVs that have a minimum diameter of 2.5 μm and a pitch of 6.3 μm with a line of 2 μm and space of 0.64 μm. In total they have over 35,000 TSVs – about 15,000 connecting the pixel substrate and the DRAM substrate and about 20,000 more connecting the DRAM substrate to the logic substrate.
The chip achieved 120 fps for all 19.3M pixels and can produce 960 fps FHD (1,920 x 1,080) super slow motion video."
 |
| 3-layer stacking process flow |
 |
| TEM cross-section |
Thursday, February 08, 2018
Velodyne Talks about LiDAR Advantages, Tesla Denies the Need
Velodyne publishes a video on LiDAR advantages in automotive applications:
SeekingAlpha publishes Tesla Q4 2017 earnings call transcript with CEO Elon Musk saying:
Q: "Elon, on your autonomous vehicle strategy, why do you believe that your current hardware set of only camera plus radar is going to be able to get you to fully-validated autonomous vehicle system? Most of your competitors noted that they need redundancy from lidar hardware to given the robustness of the 3D point cloud and the data that's generated. What are they missing in their software stack and their algorithms that Tesla is able to obtain from just the camera and plus radar?
Further, what would be your response if the regulatory bodies required that level of redundancy is really needed from an incremental lidar hardware?"
Elon Musk: "Well, first of all, I should say there's actually three sensor systems. There are cameras, redundant forward cameras, there's the forward radar, and there are the ultrasonics for near field. So, the third is also – the third set is also important for near-field stuff, just as it is for human.
But I think it's pretty obvious that the road system is geared towards passive optical. We have to solve passive optical image recognition, extremely well in order to be able to drive in any given environment and the changing environment. We must solve passive optical image recognition. We must solve it extremely well.
At the point at which you have solved it extremely well, what is the point in having active optical, meaning lidar, which does not – which cannot read signs; it's just giving you – in my view, it is a crutch that will drive companies to a local maximum that they will find very difficult to get out of.
If you take the hard path of a sophisticated neural net that's capable of advanced image recognition, then I think you achieve the goal maximum. And you combine that with increasingly sophisticated radar and if you're going to pick active photon generator, doing so in 400 nanometer to 700 nanometer wavelength is pretty silly, since you're getting that passively.
You would want to do active photon generation in the radar frequencies of approximately around 4 millimeters because that is occlusion penetrating. And you can essentially see through snow, rain, dust, fog, anything. So, it's just I find it quite puzzling that companies would choose to do an active photon system in the wrong wavelength. They're going to have a whole bunch of expensive equipment, most of which makes the car expensive, ugly and unnecessary. And I think they will find themselves at a competitive disadvantage.
Now perhaps I am wrong. In which case, I'll look like a fool. But I am quite certain that I am not."
SeekingAlpha publishes Tesla Q4 2017 earnings call transcript with CEO Elon Musk saying:
Q: "Elon, on your autonomous vehicle strategy, why do you believe that your current hardware set of only camera plus radar is going to be able to get you to fully-validated autonomous vehicle system? Most of your competitors noted that they need redundancy from lidar hardware to given the robustness of the 3D point cloud and the data that's generated. What are they missing in their software stack and their algorithms that Tesla is able to obtain from just the camera and plus radar?
Further, what would be your response if the regulatory bodies required that level of redundancy is really needed from an incremental lidar hardware?"
Elon Musk: "Well, first of all, I should say there's actually three sensor systems. There are cameras, redundant forward cameras, there's the forward radar, and there are the ultrasonics for near field. So, the third is also – the third set is also important for near-field stuff, just as it is for human.
But I think it's pretty obvious that the road system is geared towards passive optical. We have to solve passive optical image recognition, extremely well in order to be able to drive in any given environment and the changing environment. We must solve passive optical image recognition. We must solve it extremely well.
At the point at which you have solved it extremely well, what is the point in having active optical, meaning lidar, which does not – which cannot read signs; it's just giving you – in my view, it is a crutch that will drive companies to a local maximum that they will find very difficult to get out of.
If you take the hard path of a sophisticated neural net that's capable of advanced image recognition, then I think you achieve the goal maximum. And you combine that with increasingly sophisticated radar and if you're going to pick active photon generator, doing so in 400 nanometer to 700 nanometer wavelength is pretty silly, since you're getting that passively.
You would want to do active photon generation in the radar frequencies of approximately around 4 millimeters because that is occlusion penetrating. And you can essentially see through snow, rain, dust, fog, anything. So, it's just I find it quite puzzling that companies would choose to do an active photon system in the wrong wavelength. They're going to have a whole bunch of expensive equipment, most of which makes the car expensive, ugly and unnecessary. And I think they will find themselves at a competitive disadvantage.
Now perhaps I am wrong. In which case, I'll look like a fool. But I am quite certain that I am not."
Subscribe to:
Posts (Atom)