This paper presents the image quality issues encountered at system level when working with infrared imagers. This work highlights some characteristics that are usually not specified for off-theshelf components. The first part of this work discusses the need for short-term and long-term stability. We present a comparative study of two sensors in the MW band based on the residual fixed pattern noise and defective pixels. In the second part we present the issues of patterns defects. We conduct an experimental study on twelve persons to estimate the perception threshold of such defects within an infrared sequence.
A thirty months ESA project started in March 2008, whose purpose is to expand and assess the performance of broadband (11-15μm) quantum detectors for spectro-imaging applications: Fourier Transform Spectrometers and Dispersive Spectrometers. We present here the technical requirements, the development approach chosen as well as preliminary signal to noise ratio (SNR) calculations. Our approach is fully compatible with the final array format (1024x256, pitch 50-60μm). We expect the requested uniformity, operability and SNR levels to be achieved at the goal temperatures (60K for FTS applications and 50K for DS applications). The performance level will be demonstrated on 256x256, 50μm pitch arrays. Also, operability and uniformity issues will be addressed on large mechanical 1024x256 hybrid arrays.
Over the past years, a huge interest has grown in both the scientific and the industrial communities for miniaturized and functionalized cameras featuring new capabilities such as depth estimation or multispectral imaging. As a consequence, new optical architectures such as the plenoptic camera have been proposed and studied, primarily in the visible spectrum. These cameras usually include an optical element such as a microlens array or a prism array in order to obtain multiple sub-images of a same object point on the sensor, allowing for single snapshot image refocusing or depth estimation. In the meantime, recent developments in cooled infrared focal plane arrays technology have led to smaller pixel pitch and bigger formats, thus slowly reducing the resolution gap that existed between visible and infrared cameras. This gain in resolution enables the design of more functionalized infrared imaging systems, thus answering the critical need for more features in a limited volume, especially for military applications. However, the use of cooled infrared sensors brings an additional challenge to the design of such cameras because of the specific assembly in which the sensor has to be embedded called a dewar. In this paper we explain how we overcame these constraints to design and implement three different cooled infrared cameras with single focal plane array depth estimation capabilities. We then evaluate the performance of these cameras in terms of range and precision of the depth estimation and conclude on their potential applications.
A thirty months ESA project started in March 2008, whose overall purpose is to expand and assess the
performance of broadband (11-15 µm) quantum detectors for spectro-imaging applications: Dispersive Spectrometers
(DS) and Fourier Transform Spectrometers (FTS). We present here the technical requirements, the development
approach chosen as well as preliminary signal to noise ratio (SNR) calculations. Our approach is fully compatible with
the final array format (1024x256, pitch 50-60μm). We expect the requested uniformity, operability and SNR levels to be
achieved at the goal temperatures (60K for FTS applications and 50K for DS applications). The performance level will
be demonstrated on 256x256, 50µm pitch arrays. Also, operability and uniformity issues will be addressed on large
mechanical 1024x256 hybrid arrays.
Since 2005, the THALES Group has successfully manufactured TV/4 format QWIP sensitive arrays in high rate
production at the THALES Research and Technology Laboratory. The full-TV array manufacturing started in 2007.
Uniformity and stability were the key parameters which led to the selection of this technology for thermal cameras.
Another widely claimed advantage for QWIPs was the versatility of the band-gap engineering and of the III-V
processing allowing the custom design of quantum structures to fulfill the requirements of specific applications such as:
very long wavelength (VLWIR); multi-spectral detection; and polarimetric detection.
Serial production of CATHERINE-XP and CATHERINE-MP has now started for the various programs for which both
cameras have been selected. A review of the QWIP Production status, CATHERINE achievements and current programs
are presented. THALES has based its current strategy on very compact TI in order to address the largest range of
platforms and applications, and is working in cooperation with Sofradir and TRT / III-Vlab on the evolution of the
product to take advantage of the new capabilities offered by QWIP technology. In addition, future products based on
dual band, multi-band and polarimetric imagery are under development. An overview of these developments is presented.
Standard GaAs/AlGaAs Quantum Well Infrared Photodetectors (QWIP) are now seriously considered as a technological choice for the 3<sup>rd</sup> generation of thermal imagers.
Since 2001, the THALES Group has been manufacturing sensitive arrays using QWIP technology based on AsGa techniques through THALES Research and Technology Laboratory. This QWIP technology allows the realisation of large staring arrays for Thermal Imagers (TI) working in the Infrared region of the spectrum. A review of the current QWIP products is presented (LWIR, MWIR and dual color FPAs).
The main advantage of this GaAs detector technology is that it is also used for other commercial devices. The duality of this QWIP technology has lead to important improvements over the last ten years and it reaches now an undeniable level of maturity. As a result, the processing of large substrate and a good characteristic uniformity, which are the key parameters for reaching high production yield, are already achieved. Concerning the defective pixels, the main common features are a high operability (above 99.9%) and a low number of clusters including a maximum of 5 dead pixels.
Another advantage of this III-V technology is the versatility of the design and processing phases. It allows customizing both the quantum structure and the pixel architecture in order to fulfill the requirements of any specific applications. The spectral response of QWIPs is intrinsically resonant but the quantum structure can be designed for a given detection wavelength window ranging from MWIR, LWIR to VLWIR.
Some parameters of integration of a Quantum Cascade Detector (QCD) in an infrared imaging system are studied. Performances of QCD are first presented : absorption and responsivity spectra, peak responsivity (around 44 mA/W), resistivity at zero bias and detectivity. Quantum efficiency and photoconduction gain are deduced from these results. Finally the consequences of an integration of such a detector in a readout circuit are studied in terms of saturation of an external capacitor.
Successful past experience of implementing long wave MCT 1st and 2nd Generation thermal imagers has demonstrated to THALES Optronics that MCT presents difficult challenges when correcting non-uniformity errors caused by rapidly changing detector element gain and offset drifts. These problems become even more demanding when the move is made from long linear arrays to focal plane arrays due to the significantly larger number of detector elements. Relaxation of these demands would make a significant impact on the price/performance trade which inevitably occurs in a camera development. In recognition of the need to offer UK MOD best value, THALES Optronics has initiated a programme to achieve a SXGA resolution camera and is working with UK MOD, over a two year period, to investigate whether an alternative technology can maintain the high resolution required whilst achieving a downward step change in price. The selected technology is 3rd Generation Gallium Arsenide long wave Quantum Well Infra-red Photodiode (QWIP) chosen because initial indications are that drift rates are orders of magnitude slower than MCT. The programme involves studies to determine effects of defect clusters, bimodalism, non-uniformity correction levels and higher than normal operating temperatures on achieving acceptable performance, including logistics, in user scenarios whilst maximising detector yield. Development of demonstrator IR camera hardware (technology readiness level 6/7) based on a THALES Research & Technology QWIP array is also part of the programme.
On standard CMOS processes, basically two photosensors may be designed: photodiodes or vertical bipolar phototransistors. A trade-off must be found between the area of the sensor, its sensitivity and its bandwidth. In most designs, the high sensitivity of the sensor is a key point and led to choosing a phototransistor based solution. However this choice is made at the expense of the bandwidth of the sensor. For small currents, an analysis shows that it is mainly proportional to the base-emitter capacitance C<SUB>be</SUB> and to the collector current. Hence, in the case of a floating base bipolar and for a given current, the only way of reducing C<SUB>be</SUB> is to decrease the emitter area. On the other hand, the sensitivity is to be preserved. We have proposed and tested an original sensor based on the splitting of phototransistors. The basic idea is to use minimum size emitter bipolar transistors and to increase their collector-base junction perimeter. Thanks to this design, for a given sensor area, the bandwidth has been improved by a factor of 3 and the sensitivity has been preserved. This solution has been successfully used on an operational retina performing stochastic computations at video rates. In particular, thanks to our design, we have been able to successfully implement a 150 by 50 micrometer<SUP>2</SUP> optoelectronic random generator providing up to 100,000 random variables per second.
An optoelectronic device for generating random numbers at high flow rates is presented. Such a device can be exploited for implementating massive arrays of stochastic decision-making elements encountered in simulated annealing and in Boltzmann machine algorithms. It is comprised of an optical random number generator and an array of electronic neuron elements. The random number arrays are derived from the continually changing modal noise of a step-index optical fiber, and the neuron elements include photodiodes and a few associated electronic components. To demonstrate the feasibility of large arrays of neuron elements, a circuit with three neurons has been implemented and tested. The results reveal that 10<sup>4</sup> neuron elements can be integrated into 1 cm<sup>2</sup> and that flow rates of 10<sup>10</sup> random numbers per second are possible.