Imaging system having multiple image copying and hierarchical busing

Claim 2

An imaging system wherein video data from each video frame of video data words is multiplexed serially by a row, including means for enhancing said video data by operating on successive kernels of said data words, comprising: a plurality of main processors and memories individually associated with said processors, each of said processors corresponding to an individually assigned block of video data words in said frame; a plurality of bus control means, individually associated with said processors, each one of said bus control means for selecting those data words from said video data words which lie within the corresponding assigned data word block; data bus means for simultaneously distributing each of said data words to said k bus control means; and a plurality of internal data bus means individually associated with said plurality of bus control means, each of said internal data bus means for carrying the selected ones of said data words from the corresponding bus control means to the corresponding one of said memories and for permitting the corresponding one of said processors to access data words in said corresponding memory, wherein each processor is programmed with selected operator means for generating an enhanced value from each kernel of data words stored in the corresponding memory.

Claim 3

The imaging system of claim 1 further comprising host computer means for programming said main processors, wherein a different image processing algorithm may be substituted into said main processors.

Claim 4

The imaging system of claim 1 wherein each of said bus control means comprise a sub-processor.

Claim 5

The imaging system of claim 3 wherein said host computer means for programming programs said subprocessor.

Claim 6

A video imager having the capability to perform programmable image enhancement in real time, comprising: a source of video data words representing an array of columns and rows of data words, said source being organized serially by rows; a plurality of r main processors, each programmed to operate on a kernel comprising a specific number of said columns and rows of said data words, each kernel being disposed symmetrically with respect to an operand one of said data words and each of said processors corresponding to an individually assigned block of data words in said video source; a plurality of memories, each utilized by one of said processors; a plurality of bus control means each individually associated with one of said processors, each one of said bus control means for selecting those data words from said video source which lie within the data word block corresponding to one of said processors; data bus means for simultaneously distributing each of said data words to said bus control means; and a plurality of internal data bus means, each one of said internal data bus means for carrying the selected ones of said data words from the corresponding bus control means to the corresponding one of said memories and for permitting the corresponding one of said processors to access data words in said one memory, wherein each processor is programmed with selected operating means for generating an enhanced value from each kernel stored in the corresponding memory.

Claim 7

A video imager having the capability to perform programmable image enhancement in real time, comprising: a source of video data words representing an array of columns and rows of data words, said source being organized serially by rows; a plurality of r main processors, each programmed to operate on a kernel comprising a specific number of said columns and rows of said data words, each kernel being disposed symmetrically with respect to an operand one of said data words and each of said processors corresponding to an individually assigned block of data words in said video source; a plurality of memories, each utilized by one of said processors; a plurality of bus control means each individually associated with one of said processors, each one of said bus control means for selecting those data words from said video source which lie within the data word block corresponding to one of said processors; data bus means for simultaneously distributing each of said data words to said bus control means; and a plurality of internal data bus means, each one of said internal data bus means for carrying the selected ones of said data words from the corresponding bus control means to the corresponding one of said memories and for permitting the corresponding one of said processors to access data words in said one memory, wherein each processor is programmed with selected operating means for generating an enhanced value from each kernel stored in the corresponding memory.

Claim 8

The device of claim 5, further comprising: a plurality of internal address bus means, each one of said internal address bus means for carrying the address words corresponding to the selected data words from one of said bus control means to the corresponding memory and for permitting the corresponding processor to access address words in said corresponding memory.

Claim 9

The imager of claim 6 wherein the jth one of said bus control means comprises: a data register connected to said data bus means; an address register connected to said address bus means; an internal sub-processor connected to said data and address registers and connected to said internal data bus means, said internal sub-processor comprising: means for determining whether the contents k of said address register lies within a predetermined address range; and means for fetching the contents of said data register and putting it on said internal bus means in response to said determining means.

Claim 10

The device of claim 7 wherein said internal sub-processors are programmable to permit a selection of the size of the kernel operated upon by said main processors.

Claim 11

The device of claim 7 wherein said host computer means for programming programs said internal sub-processors.

Claim 12

The device of claim 9 wherein said first buffer bus control means comprises a buffer processor.

Claim 13

The device of claim 10 wherein said host computer means for programming programs said buffer processor.

Claim 14

The device of claim 5 wherein said data bus means has a first pixel processing data rate corresponding to the data rate of said video data word source, and said internal bus means has a second pixel processing data rate reduced from said first data rate by an amount substantially proportional to the number of said bus control means, said second pixel processing data rate being compatible with the computation rate of said processors.

Claim 15

The device of claim 5 wherein said data bus means further comprises a plurality of s buffers, each of said buffers controlling transmission between said video data word source and a fractional plurality of said bus control means, each of said buffers comprising: a buffer memory; a first buffer bus control means connected to said data bus means for receiving data words from said video data word source and for selecting and storing in said buffer memory those data words lying within an individually assigned block of data words in said array; an intermediate output bus connected to said fractional plurality of said bus control means; a second buffer bus control means connected between said buffer memory and said intermediate bus, said second buffer bus control means for transmitting the data words stored in said buffer memory to said intermediate bus, wherein said data bus has a first pixel processing data rate corresponding to the data rate of said video source, said intermediate data bus has a second pixel processing data rate substantially reduced from said first rate by a factor proportional to 1/s, and said internal data bus means has a pixel processing data rate further substantially reduced from said first data rate by a factor proportional to 1/(r×s).

Claim 16

The device of claim 13 wherein said fractional plurality is proportional to r/s.

Claim 17

The device of claim 5 wherein each of said main processors are programmed to perform a Sobel operation upon each of said kernels to enhance edge detection by said imager.

Claim 18

The device of claim 5 further comprising host computer means for programming said plurality of r main processors.