A Relative Imaging CMOS Image Sensor for High Dynamic Range and High Frame-Rate Machine Vision Imaging Applications

This paper proposes an unconventional image acquisition scheme for machine vision applications, based on detecting ratios of illumination (pixel) intensities. Detecting relative ratios enables capturing the scene features and patterns almost independently from the local scene illumination resulting in potentially extremely high dynamic range. Moreover, detecting signal ratios using a fully differential circuit optimally suits the intrinsic nature of very large scale integrated design. A scalable and compact hardware implementation is proposed as a proof-of-concept toward relative image acquisition. The proposed photo-current-ratio-detecting pixels completely bypass the need of conventional photo-current integration which enables high frame-rate operation of up to 24 000 frames/s. The pulse-width-modulated output of the proposed pixel is captured by compact column-parallel readout circuits based on digital counters. The developed 32 x 32 pixel array prototype CMOS image sensor consumes 4 mW of power operating at a nominal 9765 frames/s frame rate, and 6.8 mW of power operating at a maximum 24 000 frames/s. The presented prototype design is fully scalable toward newer CMOS fabrication nodes and higher sensor resolution.

Smart Imaging and On-Focal-Plane Image Acquisition and Processing in Standard CMOS Technology

Nikola Katic

In recent years, modern imaging sensors and systems have become increasingly complex following the growing demand for high-quality and high-resolution imaging. Commercially available sensors having 30-40 mega-pixel resolutions are common nowadays, while professional and scientific imaging systems (e.g. telescope and satellite sensors) often have hundreds of mega-pixels or even giga-pixel resolution. Such sensors or multi-sensor systems generate extreme amounts of data and require an excessive amount of power in order to process and stream-out the acquired data at the required speed. Consequently, three difficult and closely related bottlenecks appear in such systems: power consumption, data bandwidth and memory storage. Nevertheless, the vast majority of applications (such as machine vision, bio-medical, surveillance, space imaging, etc.) do not process all the data acquired from the conventional image acquisition. In fact, most of the image (video) data is extremely redundant in both temporal and spatial domain. Moreover, in many of these applications, this excessive data is simply discarded despite the additional resources that were used to acquire it. This Thesis explores new and and "smart" ways of image acquisition that can exploit this redundancy and reduce the required image data. The ultimate goal of the research conducts into reducing the power consumption, system bandwidth, memory requirements and improve the overall system efficiency in numerous modern image acquisition applications. The main focus of this Thesis is the investigation of three unconventional approaches to image acquisition that offer power, bandwidth and memory storage reduction. All three methods are explored in the context of CMOS integrated circuit implementation directly on the focal-plane in order to maximize their efficiency. The first investigated method is called "relative imaging" (RI). The image sensor detects the relative ratios of the photo-currents corresponding to the neighboring pixels without performing pixel integration. This method allows high frame-rates with a very high dynamic range. Moreover, detecting the photo-current ratios (instead of the absolute amplitudes) corresponds better to the intrinsic nature of VLSI design. The second investigated method which enables the reduction of the required image information is compressive sampling (CS). Compressive sampling exploits image sparsity in order to reduce the amount of information that is required to guarantee correct image reconstruction. The third method, multi-band edge detection (MBED), decomposes the image content into separate frequency bands and performs single-bit quantization of each frequency band separately. It achieves 1-1.33 bit-per-pixel compression level, while preserving most of the image information and demonstrating robust performance in the presence of noise and other physical non-idealities. An additional focus of this Thesis is the design of unconventional analog-to-digital converter (ADC) topologies convenient for "smart" image acquisition.

EPFL2014

Low-noise high-dynamic-range single-photon avalanche diodes with integrated PQAR circuit in a standard 55 nm BCD

Claudio Bruschini, Edoardo Charbon, Francesco Gramuglia, Pouyan Keshavarzian, Ekin Kizilkan, Myung Jae Lee

Single-photon avalanche diode (SPAD) based sensors and systems enable a variety of applications in biomedical, automotive, consumer, and security domains. While several established standard technologies, which can facilitate the design of SPAD-based systems are already in existence, challenges remain for the development of deep sub-micron monolithic integration of circuits and SPADs. In this work, we present SPADs along with pixel circuits in a standard GF 55 nm BCDL process. Two different designs demonstrate the flexibility allowed by the technology for a variety of applications. Both shallow and deep junction SPADs present excellent noise performance of less than 1 cps/mu m(2) at 3 V excess bias. An integrated passive-quench active-recharge (PQAR) circuit is used in conjunction with the SPADs, which enables a dead time of less than 2 ns, easily allowing for high dynamic range applications that require > 100 Mcps such as quantum communication and information technologies. The deep and shallow junction SPADs demonstrate an afterpulsing probability of < 0.5 % and < 2 % at 3V excess bias, respectively. The dead time is adjustable through analog control of the active-recharge circuit, allowing for afterpulsing reduction to below 0.1 % while maintaining Mcps operation. The shallow junction, which has a breakdown voltage of about 18 V and a peak sensitivity at 430 nm is particularly interesting for applications requiring low supply voltage, whereas the deep SPAD, which demonstrates > 4 % photon detection probability (PDP) at 940 nm, can be implemented in LiDAR sensors that require enhanced sensitivity at near-infrared (NIR) wavelengths. The measured timing jitter of both SPADs is < 50 ps FWHM at 3 V excess bias and 780 nm.

SPIE-INT SOC OPTICAL ENGINEERING2022

Smart Imaging and On-Focal-Plane Image Acquisition and Processing in Standard CMOS Technology

Nikola Katic

EPFL2014

Low-noise high-dynamic-range single-photon avalanche diodes with integrated PQAR circuit in a standard 55 nm BCD

Claudio Bruschini, Edoardo Charbon, Francesco Gramuglia, Pouyan Keshavarzian, Ekin Kizilkan, Myung Jae Lee

SPIE-INT SOC OPTICAL ENGINEERING2022