Time-of-Flight Range Image Sensor (TRIS)

PRIN2007-TRIS DIT-PRJ-08-045

Status NOT active project
DISI role Coordinator
Project type Research Project
Dimension National
Acquisition date 2008-07-23
Start date 2008-09-22
End date 2010-09-22

Project details

Project astract Optical sensors for three-dimensional vision (3D imagers) are able to yield the spatial coordinates (x,y,z) relevant to the surface of the objects present in a scene [1,2] and candidate themselves to replace standard 2D solid-state imagers in many applications (virtual reality, automatic navigation, medicine, security, etc.). Among the optical techniques for range measurement, the Time-Of-Flight (TOF) is particularly important: it is based on the measurement, either direct or indirect, of the time<br/>needed for an optical signal to travel from a source to a target and back to a sensor. TOF techniques provides the best performance in terms of acquisition speed, reliability and overall cost of the system, and are most suited to integrated implementation of the measurement system. <br/><br/>The main goal of the TRIS project is the realization of a TOF-based, 3D image sensor with a standard CMOS technology, having better performance with respect to the state-of-the-art in terms of the minimum pixel size (that is currently 15um in the best case) and/or of the maximum modulation frequency (currently limited to 20MHz for most of the existing systems). To this purpose, the project intends to use a novel type of photodetector (current assisted photo-demodulator) able to operate<br/>at high frequency and to allow for a significant simplification of the read-out circuits at the pixel level. It is planned to perform: (i) a thorough theoretical study of the distance measurement system based on TOF; (ii) design/realization/characterization of all the building blocks of the system (single photodemodulators, main read-out and control circuits, illumination modules, microlenses arrays for fill factor recovery); (iii) design/realization/characterization of the CMOS 3D image<br/>sensor, and realization of a prototype demonstrator interfaced with a PC and able to acquire/visualize/process 2D+3D images. <br/><br/>The proposed 3D image sensor is extremely innovative, since it will feature a 64x64 pixel array (i.e., a 2D resolution high enough for many applications) with a 10-15 μm pitch (i.e., the best result ever achieved), and the capability to measure the distance in real time in a range up to 6 - 10 meters with an accuracy of few centimeters. The significant competence and experience of the proponents in the fields of electro-optical instrumentation, CMOS image sensor design and device simulation/modeling represent an ideal basis to successfully target the project objectives and to accomplish excellent scientific results with interesting outcomes for applications.
Keywords CMOS IMAGE SENSORS, RANGE OPTICAL MEASUREMENT, TIME OF FLIGHT, ACTIVE PIXEL SENSOR, CURRENT ASSISTED PHOTO DEMODULATORS
Fundings 183173 €
Partners
  • University of Pavia
  • University of Modena and Reggio Emilia
  • DIT - UniTN

DISI Sub-project details

Project astract Besides coordinating the overall project, the RU-I of Trento will work at the design of CMOS image sensors and at their electro-optical and functional characterization,tightly interacting with the other research units as far as the system-related (RU-II) and device-related (RU-III) tasks are concerned.<br/>The specific activities to be carried out by the RU-I are summarized in the following. The critical aspects for circuit design are the distribution to the pixel matrix of the high-frequency modulation signals, and the read-out circuits to be integrated at the pixel level, that should be specifically tailored to the photodemodulator characteristics. The first CMOS chip is indeed aimed at the design and characterization of the critical building blocks performing the above mentioned functions:<br/>among them, in particular, the voltage buffer able to feed the modulating signals to the pixel matrix and the entire read-out channel, including the in-pixel amplifier, that should ideally remove the common mode components of the signals, the column amplifier, the Correlated Double Sampling filters for noise reduction and the output buffer. On the basis of the experimental results obtained from the first chip and of further simulations, the design of the in-pixel amplifier will be optimized, the<br/>design of the "peripheral" electronics will be reviewed and optimized, and the address and control circuits will be designed, relying on existing solutions that the proponents have successfully adopted in the past few years for the design of CMOS image sensors. The second CMOS chip will include a complete 3D image sensor, featuring a 64x64 active pixel matrix, and will be thoroughly characterized: all the main electro-optical parameters will be measured as a function of the optical<br/>power and of the wavelength, and the noise parameters (fixed pattern noise and temporal noise) will also be measured; finally, a demonstrator will be set-up, having a LABVIEW-based interface with a PC, and several 2D+3D images will be acquired in different illumination conditions.<br/><br/>Moreover, the RU-I will interact with the other research units by performing the following activities:<br/>a) as for the photodemodulator: (i) the electro-optical characterization of photodemodulator test structures possibly available from previous runs, and intended to provided a feedback to simulations, will be performed; (ii) a device behavioral macromodel will be implemented within the CADENCE software to allow for the photodemodulator simulation at the circuit level; (iii) test structures for the extraction of the main physical and technological parameters of interest for the photodemodulator operation will de designed; (iv) the RU-I will contribute to the design of the photodemodulators, featuring different topological, geometrical and process options, and to their electro-optical characterization.b) as for the theoretical analysis of the system, the RU-I will contribute to the definition of the overall specifications and to the projections of the same on the<br/>performance of all the components involved. Moreover, the geometrical characteristics of the pixels (pixel pitch and photodemodulator area) will be made available to RU-II for the design and realization of the microlenses arrays for fill-factor recovery. The latter will be assembled with the CMOS image sensors, and a comparative evaluation of the system performance with/without microlenses arrays will be carried out jointly by RU-I and RU-II.
Keywords CMOS IMAGE SENSORS, RANGE OPTICAL MEASUREMENT, TIME OF FLIGHT, ACTIVE PIXEL SENSOR, CURRENT ASSISTED PHOTO DEMODULATORS
Fundings 75000 €
Manager Gian Franco Dalla Betta
Participating RP