Abstract: In a SIMD architecture having a two dimensional array of processing elements, where a controller broadcasts instructions to all processing elements in the array, a dynamically reconfigurable switching means useful to connect four of the processing elements in the array into a group in accordance with either the broadcast instruction of the controller or a special communication instruction held in one processing element of the group, the switch includes at least one dataline connected to each processing element in the group. A multiplexer is connected to each data line and to the controller and to a configuration register. It is adapted to load the special communication instruction from the one processing element in the group into a configuration register and to operate in accord with either the broadcast instruction from the controller or the contents of the configuration register to select one of the four data lines as a source of data and applying the data therefrom to a source output port.

Abstract: A method of detection and tracking of targets using multiple angle-only sensors is disclosed. The method is computationally efficient and is capable of detecting and tracking multiple targets in true three dimensional space. The method comprises distributing a set of modules in space and associating those modules with bearing lines from each sensor to each target. Through repetitive iterative associations between modules and bearing lines, the modules converge on the targets present.

Abstract: A multiple elastic feature network having N independent sets of M feature specific neurons that respond to the temporal properties of a number of targets. The targets are represented by sets of M feature specific coordinates. Each of the neurons and its coordinates are labeled with a different one of the feature types, and have an associated receptive field and distortion, locking and expectation parameters. A sequence of candidate coordinates, which include the targets' coordinates as well as false or ghost coordinates, is input to the MEFN. Each successive candidate coordinate selects the closest neuron that has the same feature type and whose receptive field includes the candidate coordinates. The coordinates of the selected neuron and the other neurons in its set are adjusted towards the candidate coordinates.

Abstract: A multiple elastic feature network having N independent sets of M feature specific neurons that respond to the temporal properties of a number of targets. The targets are represented by sets of M feature specific coordinates. Each of the neurons and its coordinates are labeled with a different one of the feature types, and have an associated receptive field and distortion, locking and expectation parameters. A sequence of candidate coordinates, which include the targets' coordinates as well as false or ghost coordinates, is input to the MEFN. Each successive candidate coordinate selects the closest neuron that has the same feature type and whose receptive field includes the candidate coordinates. The coordinates of the selected neuron and the other neurons in its set are adjusted towards the candidate coordinates.

Abstract: In a parallel processing architecture following the Single Instruction stream Multiple Data stream execution paradigm where a controller element is connected to at least one processing element with a local memory having a local memory address shift register adapted to receive and retain therein a globally broadcast memory base register address value received from the controller element for use by the processing element for access and transfer of data between the processing element and its respective local memory, a computer architecture for implementing indirect addressing and look-up tables includes a processing element shift register associated with the at least one processing element and adapted to receive and retain therein a local memory offset address value calculated or loaded by the associated processing element in accord with a first predetermined set of instructions.