Massimo Gottardi

Novel CMOS image sensor with a 132-dB dynamic range

A CMOS image sensor providing an ultrawide dynamic range with a piecewise linear response is presented. The active pixel is based on a novel architecture which implements a voltage comparator and an analog memory to detect and store the information on the integration time needed to reach saturation, while also maintaining the standard integrated photo-current signal. A 128/spl times/64 pixel array has been designed and fabricated in 0.35-/spl mu/m, 3.3-V CMOS technology. The chip measures 2.67/spl times/4.90 mm/sup 2/ with a pixel size of 24.65/spl times/24.65 /spl mu/m/sup 2/ and a fill factor of about 11%. The sensor has been fully characterized and the measured dynamic range turned out to be 132 dB with a power consumption of 14 mW at video frame rate. The sensor features also good noise performance with a temporal noise of 0.2% (1.7%) and a fixed pattern noise of 0.4% (1.5%) at low (high) irradiance.

A 100

An ultra-low power 128 times 64 pixels vision sensor is here presented, featuring pixel-level spatial contrast extraction and binarization. The asynchronous readout only dispatches the addresses of the asserted pixels in bursts of 80 MB/s, significantly reducing the amount of data at the output. The pixel-embedded binary frame buffer allows the sensor to directly process visual information, such as motion and background subtraction, which are the most useful filters in machine vision applications. The presented sensor consumes less than 100 muW at 50 fps with 25% of pixel activity. Power consumption can be further reduced down to about 30 muW by operating the sensor in Idle-Mode, thus minimizing the sensor activity at the ouput.

\mu

Fast collection of three-dimensional (3D) data sets is required in a growing number of applications like robotic guidance, security and automotive. Scannerless time-of-flight (TOF) based active 3D vision systems are capable of collecting depth profiles of entire scenes in fractions of seconds but have the disadvantage of using high voltage and expensive electronic components. In this contribution we describe the design and test of an integrated smart pixel with 3D vision capability, which attempts to address this kind of problem. The pixel, fabricated in standard 0.6 /spl mu/m CMOS technology, is suitable to be used in scannerless ranging systems. By means of a pulsed laser source, distance estimation is obtained by integrating the back-reflected signal in successive time windows.

W 128

A 64 × 64-pixel ultra-low power vision sensor is presented, performing pixel-level dynamic background subtraction as the low-level processing layer of an algorithm for scene interpretation. The pixel embeds two digitally-programmable Switched-Capacitors Low-Pass Filters (SC-LPF) and two clocked comparators, aimed at detecting any anomalous behavior of the current photo-generated signal with respect to its past history. The 45 T, 26 μm square pixel has a fill-factor of 12%. The vision sensor has been fabricated in a 0.35 μm 2P3M CMOS process, powered with 3.3 V, and consumes 33 μ W at 13 fps, which corresponds to 620 pW/frame.pixel.

\times

A prototype of a 34 /spl times/ 34 pixel image sensor, implementing real-time analog image processing, is presented. Edge detection, motion detection, image amplification, and dynamic-range boosting are executed at pixel level by means of a highly interconnected pixel architecture based on the absolute value of the difference among neighbor pixels. The analog operations are performed over a kernel of 3 /spl times/ 3 pixels. The square pixel, consisting of 30 transistors, has a pitch of 35 /spl mu/m with a fill-factor of 20%. The chip was fabricated in a 0.35 /spl mu/m CMOS technology, and its power consumption is 6 mW with 3.3 V power supply. The device was fully characterized and achieves a dynamic range of 50 dB with a light power density of 150 nW/mm/sup 2/ and a frame rate of 30 frame/s. The measured fixed pattern noise corresponds to 1.1% of the saturation level. The sensors dynamic range can be extended up to 96 dB using the double-sampling technique.

64 Pixels Contrast-Based Asynchronous Binary Vision Sensor for Sensor Networks Applications

A CMOS smart pixel aimed at three-dimensional vision applications is introduced. It is suitable for scannerless laser ranging systems which employ the indirect time-of-flight measuring technique to recover distance information. The pixel is operated with trains of light pulses generated by an external source to illuminate the scene and contains most of the processing electronics to perform signal accumulation and noise reduction operations. The smart pixel architecture includes an N-well photodiode plus a self-biasing voltage amplifier and a switched-capacitor fully differential stage. The pixel is fabricated in standard CMOS 0.6 /spl mu/m technology and measures 180/spl times/160 /spl mu/m/sup 2/ (including the photodiode) with a fill factor of 14%. Electrooptical test results confirm the smart pixel functionality in a range of distance from 3 m to 9 m, and the accuracy achieved for preliminary distance measurements is 15 cm. Both the accuracy and the extension of the range of distance are supposed to be improved by reducing setup and environmental noise contributions that limit the pixel performance.

A CMOS smart pixel for active 3D vision applications

This paper presents the design principles underlying the video nodes of long-lifetime wireless networks. The hardware and firmware architectures of the system are described in detail, along with the system-power-consumption model. A prototype is introduced to validate the proposed approach. The system mounts a Flash-based field-programmable gate array and a high-dynamic-range complementary metal-oxide-semiconductor custom vision sensor. Accurate power measurements show that the overall consumption is 4.2 mW at 3.3 V in the worst case, thus achieving an improvement of two orders of magnitude with respect to video nodes for similar applications recently proposed in the literature. Powered with a 2200-mAh 3.3-V battery, the system will exhibit a typical lifetime of about three months.

A 33

A 128 times 64 pixel programmable vision sensor performs real-time analog image processing over high dynamic range images is reported. The pixel-parallel single instruction multiple data (SIMD) architecture executes real-time spatio-temporal filtering with 2.8 GOPS/mm2 and large flexibility in coefficient assignment. The sensor uses time-based and pulse-based operating modalities to execute spatio-temporal filtering on images with dynamic range up to about 100 dB. The in-pixel processing is based on two operations: the absolute value of voltage difference and accumulation of partial results. Feature extraction from the entire image is also possible without the need for image dispatching, thus optimizing both processing speed and video bandwidth. The 32.6 mum square pixel, with a fill-factor of 24%, consists of two analog memories and 28 transistors. The sensor, fabricated in 0.35 mum CMOS technology, gives a fixed pattern noise (FPN) of 0.8% and power consumption of 14 mW at 3.3 V.

\mu

A CMOS interface for a piston-type MEMS capacitive microphone is presented. It performs a capacitance-to-voltage conversion by bootstrapping the sensor through a voltage pre-amplifier, feeding a third-order sigma-delta modulator. The bootstrapping performs active parasitic compensation, improving the readout sensitivity by ~12 dB. The total current consumption is 460 uA at 1.8 V-supply. The digital output achieves 80 dBA-DR, with 63 dBA peak-SNR, using 0.35 um 2P/4M CMOS technology. The paper includes electrical and acoustic measurement results for the interface.

W 64

The aim of this contribution is to report and discuss a preliminary study and rough optimization of a novel concept of MEMS device for vibration energy harvesting, based on a multi-modal dynamic behavior. The circular-shaped device features Four-Leaf Clover-like (FLC) double spring-mass cascaded systems, kept constrained to the surrounding frame by means of four straight beams. The combination of flexural bending behavior of the slender beams plus deformable parts of the petals enable to populate the desired vibration frequency range with a number of resonant modes, and improve the energy conversion capability of the micro-transducer. The harvester device, conceived for piezoelectric mechanical into electric energy conversion, is intended to sense environmental vibrations and, thereby, its geometry is optimized to have a large concentration of resonant modes in a frequency range below 5-10 kHz. The results of FEM (Finite Element Method) based analysis performed in ANSYSTM Workbench are reported, both concerning modal and harmonic response, providing important indications related to the device geometry optimization. The analysis reported in this work is limited to the sole mechanical modeling of the proposed MEMS harvester device concept. Future developments of the study will encompass the inclusion of piezoelectric conversion in the FEM simulations, in order to have indications of the actual power levels achievable with the proposed harvester concept. Furthermore, the results of the FEM studies here discussed, will be validated against experimental data, as soon as the MEMS resonator specimens, currently under fabrication, are ready for testing.