Abstract: A system for calculating the dot product of the vectors A(k) and B(k) of dimension Q written in two's complement representation. A bus has a width sufficient for communicating signals representing a plurality of bit planes A(j,Q) and B(n,Q) corresponding to the two's complement representations of A(k) and B(k), respectively. Circuitry connected to the bus calculates C(j,n)=A(j,Q).B(n,Q) for each of said bit planes. An adder sequentially determines R(n)=C(0,n)2.sup.0 +C(1,n)2.sup.1 + . . . C(N-1,n)2.sup.(N-1) -C(N,n)2.sup.N for n=0,1,2 . . . N.

Abstract: An electronic digital system for performing discrete Fourier transforms in real time. Read only memory (ROM) modules are used as look-up tables for providing inputs for multiplier stages corresponding to sequences of sample signals as well as for logic and other input converting elements of the system. Two such systems are coupled together so that their real and imaginary output signal components may be additively and subtractively combined. The entire transform output is available one cycle period after the last sample input.

Abstract: A digital circuit executes a parallel algorithm to compute the dot product of large dimensional vectors at very high speed. The circuit may be made of a plurality of cascaded 1-bit correlator chips and a plurality of ALU chips that sum the output of the correlator chips. Alternatively, a general purpose computer architecture for implementing the invention is also provided.

Abstract: An apparatus provides matched filter output characteristics for known signals embedded in noise signals that have varying noise power distributions. The system correlates applied signals with a manipulated version of the applied signals in an analog-to-analog correlator. The manipulated signals are derived by Fourier transforming the applied signals and estimating the noise power spectral density. A filter transfer function is then calculated using the noise power estimates and stored information related to the characteristics of the known signal. The transfer function is then inverse Fourier transformed to provide a filter impulse response characteristic which is correlated with the applied signals. A method of processing applied signals to provide for an enhanced signal-to-noise ratio for varying noise power distributions is also disclosed.

Abstract: An optical processing system for performing a matrix multiplication Mx where M is a matrix and x is a vector involves two two-dimensional arrays of optically transmissive elements with each element of the first array communicating light to the corresponding elmement of the second array. The optical transmittance of each row of one of the arrays is controlled according to the values of the vector x, and the optical transmittance of each element of the other array is controlled according to the corresponding matrix value m(i,j). The first array is uniformly illuminated and the amount of light transmitted through the two arrays is detected by a suitable array of photodetectors. The output of each photodetector will be proportional to the sum of the products of the vector x(i) with each value of the matrix in the corresponding column of the matrix array.

Abstract: A two-dimensional optical correlation apparatus includes a transmissive optical mask having a transmissivity pattern corresponding to a two-dimensional reference image and an input image buffer that stores a sequence of pixels corresponding to a raster-scanned input image. The input buffer includes a subimage frame corresponding to a particular segment of the raster-scanned format. As the pixels are serially shifted through the image buffer, every subimage in the input image appears at some time in the subimage frame. The subimage frame is connected to an array of optical emitters. As the input image pixels are serially shifted, the emitter array produces a succession of two-dimensional optical signals corresponding to the succession of input image subimages shafted through the subimage frame. The output of the emitter array is projected onto the transparent mask.

Abstract: The electro-optical correlator performs pattern-recognition by forming the columnar dot product of an input image with a reference image at video through-put rates. The correlator incorporates an analog-to-digital converter for digitizing the input-image data, a plurality of serially arranged delays having a delay time equivalent to one input-image data line, each delay for receiving in turn the digitized input data and for producing a parallel output signal, and a vertical column of photodiodes for receiving the respective multiple parallel delayed signals and for converting same to an optical output. The optical output is fed through a lens, reference image mask, and lens unit and is focused on a plurality of photodiodes that are connected to a time-delay-and-integrate (TDI) array. The data from the time- delay-and-integrate array are clocked through an analog-to- digital converter which produces the final system output. The time delay and integrate array may be a charge-coupled device (CCD).

Abstract: An optical computing system is provided for performing a linear multiplication operation involving a matrix with bipolar values. A plurality of beams representing the values of a N.times.1 column matrix Q illuminate an optical mask having elements disposed in a two-dimensional (M+1).times.N array. Each element of the first M rows of the mask represents the offset of the value of a corresponding element in a known M.times.N matrix by a scalar constant c. Each element of row M+1 of the mask represents the value c. An array of photodetectors collects the light passing through the mask, with the light from each row of elements detected by a respective photodetector. The outputs of the first M photodetectors represent the product of Q and M after M has been offset by c. The output of photodetector M+1 represents the scalar value produced by the multiplication of Q and a row matrix represented by row M+1 of the mask matrix.

Abstract: A technique for correlating two images, differing in size, but of identical spatial content. The technique is based on the Cauchy-Schwarz Inequality and is derived in such a way as to eliminate the requirement for division, thereby realizing practical real time implementation with presently available electronic hardware. In the embodiment disclosed, reference and data image signals require no pre-processing for the removal of mean and gain factors prior to being individually supplied to the correlator. The correlator is built around modules which can be configured serially with delay lines to adapt to virtually any image configuration. Simulation tests have shown the embodiment to be superior to prior art correlators utilizing other algorithms under conditions of noise, offset and gain differences.

Abstract: A pulse width discriminator comprises a tapped analog delay line which preferably is a charge coupled device (CCD). The taps are spaced one sample time apart along the delay line. A correlator output is obtained by summing the outputs of all taps. A pulse amplifier output is obtained by weighting and adding the signals from selected taps. The correlator output corresponds to a low-pass filter. The sample rate is at the Nyquist rate. The pulse amplifier output is a maximum for a pulse width determined by the length of the delay line (in terms of number of sample points) divided by the sample frequency. The correlator output has a maximum corresponding to the maximum of the pulse and can be used to determine the presence of a pulse. This pulse processor is also capable of handling a plurality of inputs by multiplexing them.