A layer structure, grown in a single step, and a fabrication process were developed for the monolithic integration of AlGaAs/GaAs optoelectronic smart pixels. Metal semiconductor field-effect transistors (MESFETs), light-emitting diodes (LEDs) and photodiodes (PDs) were designed in such a way that the light emitted from the LEDs can efficiently be detected by the PDs. An example is presented of a fabricated optoelectronic smart pixel: a threshold circuit consisting of a dual-photodiode differential input, an inverter and a current-balanced output containing an LED. The circuit shows a switching energy of 2 pJ. The minimum switching power is <1 nW with a contrast ratio >1000. The maximum light output of the LED is 18 mu W With an overall power dissipation of 20 mW.

Multi-junction silicon light emitting devices (Si LED's) were designed and realised by using standard 1.2 micron and 2 micron CMOS processes with a bipolar capability and with no modifications to the processes. The designs were optimised to increase the power conversion efficiency, quantum conversion efficiency, intensity of emission and also the uniformity of emission. The devices emit light of several nW per 5 to 10 mA at 4 - 30V in the 450 to 850 nm wavelength range. All the devices operated with at least one pn junction in the field emission or avalanche breakdown mode. Quantum conversion efficiencies of up to 1.5x10(-5) have been measured which is two and a half orders to three orders of magnitude higher than previously published values for light emission from Si p-n avalanching junctions. Some directional light emission characteristics were also observed. The developed devices are viable for on-chip electro-optical applications and also for high speed chip-to-environment electro-optical applications.

A planar distance and velocity sensor based on a hybrid combination of a light source and a single CMOS photodetector/analog/digital chip is presented. The specific shape of the photodetector calculates a nonlinear geometrical correction purely optically. Hence, electronic data processing is reduced to a minimum, and it can be performed in the analog domain, on the same CMOS chip where the photodetector is placed. Since the (analog) distance and velocity signals are generated simultaneously, the device is particularly useful for time-to-contact measurements.

Two different unmodified industrial CMOS processes have been used for the integration of highly interdigitated pn structures. Under forward bias these pn junctions emit narrow-band infrared light at 1160 nm with an electrical-to-optical power conversion efficiency of typically 10−4. The same junctions show broad-band visible-light emission between 450 and 800 nm in the avalanche breakdown region under reverse bias with efficiencies of the order of 10−8. This is already enough for a first few practical applications as light-emitting devices (LEDs). No satisfactory explanation for this emission efrect, fitting all the experimentally observed electro-optical and physical properties of our silicon LEDS, has been found yet.

An optoelectronic device has been developed and built that acquires and displays the range data of an object surface in space in video real time. The recovery of depth is performed with active triangulation. A galvanometer scanner system sweeps a sheet of light across the object at a video field rate of 50 Hz. High-speed signal processing is achieved through the use of a special optical sensor and hardware implementation of the simple electronic-processing steps. Fifty range maps are generated per second and converted into a European standard video signal where the depth is encoded in gray levels or color. The image resolution currently is 128 x 500 pixels with a depth accuracy of 1.5% of the depth range. The present setup uses a 500-mW diode laser for the generation of the light sheet. A 45-mm imaging lens covers a measurement volume of 93 mm x 61 mm x 63 mm at a medium distance of 250 mm from the camera, but this can easily be adapted to other dimensions.

A 3-D camera has been developed, capable of measuring 50 complete depth images per second of arbitrary, diffusely reflecting objects. The output of this real-time 3-D camera is a CCIR video signal, in which the depth is linearly coded as the local gray-value of the video image. The measurement principle is active triangulation using a solid- state laser and two galvanometer mirrors for beam and scene scanning. No complicated electronics or computers are required because the essential ‘calculations’ (centre-of-gravity of light profiles) are carried out purely optically with a special sensor consisting of wedge pixels. The use of this smart sensor makes the 3-D video camera relatively inexpensive, and the output in video-format allows the connection with any standard video framestore for direct 3-D image processing. It is concluded that this development can be of appreciable practical interest for many applications in industry.

With ‘photo-ASICs’ comprising light-sensitive structures, light-emitting devices and analog and digital circuits, complete optical metrology systems can be integrated on a single chip. We report the realization of key components of such photo-ASICs using an industrual IC CMOS process. We achieve photodiodes with an external quantum efficiency of 50–80% in the visible spectrum and position-sensitive devices (PSDs) with a spatial non-linearity of around 0.3%. We demonstrate surface-channel CCDs with a charge-transfer efficiency of 99.8% at room temperature, as well as bucket-brigade devices (BBDs) with a lower charge-transfer efficiency of 96%. Light-emitting diodes (LEDs) are realized, emitting infrared light at 1160 nm (forward biased) and broadband visible yellow light with a spectral maximum at 640 nm (reverse biased). We discuss simple applications of passive photo-ASICs, such as a centroid detector, a 3-D camera, a motion detector and a focus sensor, used to obtain a relative measure for the local focus of an optically imaged scene, for example, in a photographic camera.

Cost-effective single-chip smart photosensors (photo-ASICs) have a wide range of applications in optical metrology systems. They may be realized through the integration of light-sensitive structures together with analog and digital electronic circuitry using a standard silicon process. Different optical sensing devices to be used as elements of such photo-ASICs (application-specific integrated circuits) were fabricated in an industrial CMOS process. Various types of vertical photodiodes, a lateral photodiode, a position-sensitive device, linear surface-channel and buried-channel charge-coupled devices, and a bucker-brigade device have been designed, fabricated in multi-project wafer services at low cost, and characterized electrically and optically. The electrical and optical performance proved to be surprisingly good.<>

A new approach to the robust recognition of objects is presented. The fundamental picture primitives employed are local orientations, rather than the more traditionally used edge positions. A simple technique of feature-matching is used, based on the accumulation of evidence in binary channels (similar to the Hough transform) followed by a weighted non- linear sum of the evidence accumulators (matched filters, similar to those used in neural networks). By layering this simple feature-matcher, a hierarchical scheme is produced whose base is a binary representation of local orientations. The individual layers represent increasing levels of abstraction in the search for an object, so that the object can be arbitrarily complex. The universal algorithm presented can be implemented in less than 100 lines of a high-level programming language (e.g., Pascal). As evidenced by practical examples of various complexities, objects can be reliably and robustly identified in a wide variety of surroundings.

A common problem in optical metrology is the determination of the exact location of an edge (a black/white transition). The use of cameras for this task has been restricted in the past because of their limited number of pixels and the lack of methods for subpixel accuracy edge detection. Analysis of the optical and electronic parts of modern solid-state cameras shows that it is possible to determine the exact location of an edge to subpixel accuracy, independently of the system's modulation transfer function. A novel algorithm for this purpose is presented together with an expression for the precision of the edge location as a function of pixel noise and edge step height. Experimental verification was carried out using a modified CCD camera coupled to an intelligent framestore (smart camera). Under optimum conditions the measured accuracy for the edge position was better than 1/140 of the pixel period, corresponding to less than 120 nm on the sensor surface of the camera. Applications of this novel method in metrology and micrometrology are discussed.