Abstract: In an associative processor particularly useful for tomographic image reconstruction, the associative memory is partitioned into an array of associative processors concurrently performing the same function with different operands on disjoint data sets. The array is dynamically reconfigurable, and the partitioning is completely flexible, with individual processors free to assume any size and topology.While the new associative processor is of broad application, its power is specifically demonstrated for tomographic image reconstruction by the well-known convolution-back-projection method.

Abstract: An associative memory cell includes a storage device having a pair of biasing nodes and a pair of output nodes, a plurality of access lines including a pair of BIT-lines and a TAG-line, and a plurality of gates controlling the accessing of the storage device by the access lines. The latter gates include a pair of sampling gates under the control of the BIT-lines and connecting the output nodes of the storage device to a pair of sampling nodes, and a pair of TAG-gates under the control of the sampling nodes and connecting the TAG-lines to one of the pair of the storage device biasing nodes.The described memory cell further includes READ-control and WRITE-control lines with the associated gates to perform the READ and WRITE functions. The pair of BIT-lines, the WRITE-control line and the READ-control line are all connected to the gate terminals of their respective gates, thereby making the device easier to drive, isolating the memory cell from the BIT-line buses, and improving reliability and speed.

Abstract: Apparatus for on-line data enhancement including apparatus for receiving data, apparatus for applying a histogram transformation to the received data including a look-up table embodied in a rewritable memory; apparatus for determining the histogram transformation in response to a histogram of the received data; and apparatus for producing a histogram of the received data.