Compressed quantized image-data transmission technique suitable for use in teleconferencing

Claim 2

In a method for transmitting in real time compressed image-data, derived from a given video signal that defines the gray-scale level value of each successive pixel of each of successive frames of uncompressed original television images that exhibit a given relatively high pixel density, over a limited bandwidth communication channel to a receiver, and for displaying at said receiver successive frames of synthesized television images derived from said transmitted compressed image-data; the improvement comprising the steps of: (a) deriving from a combination of said given video signal and a feedback signal that defines the difference in level values of corresponding pixels of at least each pair of successive frames of said original television images, an ordinally-arranged set of (N+1) separate video signals, where N is a specified positive integer, the first separate video signal of said set defining the highest spatial frequency bandpass sub-spectrum image component of the spatial frequency spectrum of a differential image corresponding to at least that pair of successive frames of said original television images, the image component of said differential image defined by said first separate video signal of said set exhibiting a first pixel density that is equal to or less than said given pixel density, each other separate video signal of said set except said last separate video signal of said set defining a spatial frequency bandpass sub-spectrum image component of said spatial frequency spectrum that is lower than that defined by its immediately preceding separate video signal of said set, the image component of said differential image defined by each other separate video signal of said set exhibiting a pixel density that is less than the pixel density exhibited by the image component defined by its immediately preceding separate video signal of said set, and said last separate video signal of said set defining a spatial frequency low pass sub-spectrum image component of said spatial frequency spectrum that is lower than that defined by its immediately preceding separate video signal of said set, the pixel density of the image component of said differential image defined by said last separate video signal of said set being no greater than the pixel density exhibited by the image component defined by its immediately preceding separate video signal of said set; (b) quantizing said video signals of said set to reduce the number of gray-scale level values of the pixels of at least one of said separate video signals of said set; (c) subsequent to said quantizing, separately compression-encoding each separate video signal of said set, and transmitting said separately compression-encoded video signals to said receiver over said limited bandwidth communication channel; and (d) subsequent to said quantizing, deriving said feedback signal from said separate video signals of said set.

Claim 3

The method defined in claim 1, wherein said given video signal is a digital signal comprised of a binary code containing a first given plurality of bits, said first given plurality of bits being sufficient in number to define with relatively high resolution the gray-scale level value of each successive pixel of each of said successive frame of said uncompressed original images; and wherein: step (b) comprises quantizing one or more of said separate video signals of said set by sufficient amounts to permit the gray-scale level value of all of the pixels of all of said separate video signals of said set, subsequent to said quantizing, to be separately compression-encoded by step (c) with a second binary code containing a second given number of bits, wherein second given number is smaller than said first given number; and step (c) comprises separately compression-encoding each separate video signal of said set, subsequent to said quantizing, with said second binary code.

Claim 4

The method defined in claim 2, wherein: the first pixel density exhibited by the image component of said differential image defined by said first separate video signal of said set derived in step (a) is equal to said given pixel density; and step (b) comprises the step of quantizing said video signals of said set to reduce the number of gray-scale level values of the pixels of said first separate video signal.

Claim 5

The method defined in claim 3, wherein: step (b) comprises the step of quantizing said video signals of said set without reducing the gray-scale level values of the pixels of said last separate video signal of said set.

Claim 6

The method defined in claim 3, wherein: step (b) includes the step of quantizing said video signal of said set to reduce the gray-scale level values of all the pixels of said first separate video signal to a single level value, the value of said single level being zero.

Claim 7

The method defined in claim 3, where N is a given plural integer, whereby said set of separate video signals includes at least one of said other video signals; and wherein: step (b) includes quantizing said separate video signals of said set to reduce the number of gray-scale level values of the pixels of said one other separate video signal.

Claim 8

The method defined in claim 6, wherein: step (b) comprises the step of quantizing said separate video signals to reduce the pixel gray scale of said first separate video signal to a first predetermined number of level values and to reduce the pixel gray scale of said one other separate video signal to a second predetermined number of level values, said second predetermined number being larger than said first predetermined number.

Claim 9

The method defined in claim 7, wherein said first given plurality of bits is eight bits, whereby the pixel gray scale of said uncompressed original images contains a maximum of 256 level values including a zero-level-value, and wherein: said second predetermined number is no greater than one-hundred-twenty eight.

Claim 10

The method defined in claim 3, wherein said first given plurality of bits is eight bits, whereby the pixel gray scale of said uncompressed original images contain a maximum of 256 level values including a zero-level-value, and wherein: said first predetermined number is no greater than one-hundred-twenty-eight.

Claim 11

In a transmitter for transmitting compressed image data to a receiver over a limited bandwidth communication channel, wherein said transmitter includes a differential pulse code modulator (DPCM), means for applying as an input to said DPCM a given video signal that defines with a given pixel density the level value of each successive pixel of each of successive frames of uncompressed two-dimensional image data, and means for data-compressing the image data derived as an output from said DPCM; the improvement in which said DPCM comprises: subtraction means having a first input to which said given video signal is applied, a second input to which a first feedback video signal is applied, and an output from which a subtraction output video signal is derived, in which the level value of each successive pixel of said subtraction output video signal is substantially equal to the difference in the respective level values of corresponding pixels of said given video signal and said first feedback video signal; a pyramid analyzer that analyzes the spatial frequency spectrum of each of successive frames of differential images represented by said subtraction output video signal applied thereto for deriving a plurality of separate sub-spectra video signals, wherein each of said sub-spectra video signals except said lowest spatial frequency sub-spectra video signal is a bandpass sub-spectrum video signal and said lowest spatial frequency sub-spectra video signal is a low-pass sub-spectrum video signal, and each bandpass sub-spectrum video signal below the highest spatial frequency bandpass sub-spectrum video signal exhibits a pixel density which is below that of its immediately next-higher bandpass sub-spectrum video signal and said low-pass sub-spectrum video signal exhibits a pixel density which is no greater than its immediately next-higher bandpass sub-spectrum video signal; means including quantizer means for separately forwarding said plurality of separate sub-spectra video signals as a plurality of separate output video signals that constitute said DPCM output, said quantizer means separately reducing the number of gray-scale level values of the pixels of at least one of the separate sub-spectra video signals prior to its being separately forwarded as an output video signal; a pyramid synthesizer responsive to said plurality of forwarded separate-output video signals applied thereto for deriving a single synthesized video signal from said plurality of forwarded output video signals, said single synthesized video signal exhibiting the same pixel density as that of said given video signal; summation means having a first input to which said single synthesized video signal is applied, a second input to which a second feedback video signal is applied, and an output from which a summation output video signal is derived, in which the level value of each successive pixel of said summation output video signal is substantially equal to the sum of the respective level values of corresponding pixels of said single synthesized video signal and said second feedback video signal; and frame delay means to which said summation output video signal is applied as an input, said frame delay means deriving said respective first and second feedback video signals as separate outputs therefrom, wherein the time delay between corresponding pixels of said given video signal and said first feedback video signal is one frame period and the time delay between corresponding pixels of said single synthesized video signal and said second feedback video signal also is one frame period.

Claim 12

In a transmitter for transmitting compressed image data to a receiver over a limited bandwidth communication channel, wherein said transmitter includes a differential pulse code modulator (DPCM), means for applying as an input to said DPCM a given video signal that defines with a given pixel density the level value of each successive pixel of each of successive frames of uncompressed two-dimensional image data, and means for data-compressing the image data derived as an output from said DPCM; the improvement in which said DPCM comprises: subtraction means having a first input to which said given video signal is applied, a second input to which a first feedback video signal is applied, and an output from which a subtraction output video signal is derived, in which the level value of each successive pixel of said subtraction output video signal is substantially equal to the difference in the respective level values of corresponding pixels of said given video signal and said first feedback video signal; a pyramid analyzer that analyzes the spatial frequency spectrum of each of successive frames of differential images represented by said subtraction output video signal applied thereto for deriving a plurality of separate sub-spectra video signals, wherein each of said sub-spectra video signals except said lowest spatial frequency sub-spectra video signal is a bandpass sub-spectrum video signal and said lowest spatial frequency sub-spectra video signal is a low-pass sub-spectrum video signal, and each bandpass sub-spectrum video signal below the highest spatial frequency bandpass sub-spectrum video signal exhibits a pixel density which is below that of its immediately next-higher bandpass sub-spectrum video signal and said low-pass sub-spectrum video signal exhibits a pixel density which is no greater than its immediately next-higher bandpass sub-spectrum video signal; means including quantizer means for separately forwarding said plurality of separate sub-spectra video signals as a plurality of separate output video signals that constitute said DPCM output, said quantizer means separately reducing the number of gray-scale level values of the pixels of at least one of the separate sub-spectra video signals prior to its being separately forwarded as an output video signal; a pyramid synthesizer responsive to said plurality of forwarded separate-output video signals applied thereto for deriving a single synthesized video signal from said plurality of forwarded output video signals, said single synthesized video signal exhibiting the same pixel density as that of said given video signal; summation means having a first input to which said single synthesized video signal is applied, a second input to which a second feedback video signal is applied, and an output from which a summation output video signal is derived, in which the level value of each successive pixel of said summation output video signal is substantially equal to the sum of the respective level values of corresponding pixels of said single synthesized video signal and said second feedback video signal; and frame delay means to which said summation output video signal is applied as an input, said frame delay means deriving said respective first and second feedback video signals as separate outputs therefrom, wherein the time delay between corresponding pixels of said given video signal and said first feedback video signal is one frame period and the time delay between corresponding pixels of said single synthesized video signal and said second feedback video signal also is one frame period.

Claim 13

The transmitter defined in claim 10, wherein: said highest spatial frequency sub-spectrum exhibits a pixel density equal to that of said given video signal, and of each bandpass sub-spectrum video signal below the highest spatial frequency bandpass sub-spectrum video signal exhibits a pixel density which is substantially one-quarter the pixel density of its immediately next-higher bandpass sub-spectrum video signal; and the nominal bandwidth of each of said bandpass sub-spectra video signals is substantially one octave of said spatial frequency spectrum in each of the two dimensions of each of said differential images.

Claim 14

The transmitter defined in claim 11, wherein: the pixel density of said low-pass sub-spectrum video signal is substantially one-quarter the pixel density of its next-higher bandpass sub-spectrum video signal.

Claim 15

The transmitter defined in claim 12, wherein: said pyramid analyzer is a Burt Pyramid analyzer; and said pyramid synthesizer is a Burt Pyramid synthesizer.

Claim 16

The transmitter defined in claim 10, wherein: said pyramid analyzer is a Burt Pyramid analyzer; and said pyramid synthesizer is a Burt Pyramid synthesizer.

Claim 17

The transmitter defined in claim 10, wherein: said quantizer means includes means for separately reducing the number of gray-scale level values of the pixels of each of said bandpass sub-spectra video signals prior to its being separately forwarded as an output video signal;

Claim 18

The transmitter defined in claim 15, wherein: said sub-spectra video signals include at least two bandpass sub-spectra video signals; and said quantizer means includes means for reducing the number of gray-scale level values of the highest spatial frequency bandpass sub-spectrum video signal to a given specified number of gray-scale level values, and separate means for reducing the number of gray-scale level values of each of the other bandpass sub-spectra video signals to a specified number of gray-scale level values associated therewith that is larger than said given specified number.

Claim 19

The transmitter defined in claim 16, wherein: the specified number of gray-scale level values associated with each other bandpass sub-spectra video signals is at least large as its immediately next-higher bandpass sub-spectrum video signal.

Claim 20

A receiver for receiving in real time compressed image data transmitted from a transmitter to said receiver over a limited bandwidth communication channel, wherein said compressed image data comprises a data-compression-encoded ordinally-arranged set of a plurality of separate video signals that respectively define substantially different sub-spectra of the spatial frequency spectrum of a differential image corresponding to at least a pair of successive frames of original television images; said receiver comprising: a data-compression decoder for separately decoding each of said separate video signals of said data-compression-encoded set to derive a decoded ordinally-arranged set of said plurality of separate video signals; and an inverse differential pulse code modulator (DPCM) for synthesizing a single output video signal defining successive frames of television images corresponding to said original television images in response to said decoded ordinally-arranged set of said plurality of separate video signals being applied as an input thereto.

Claim 21

A receiver for receiving in real time compressed image data transmitted from a transmitter to said receiver over a limited bandwidth communication channel, wherein said compressed image data comprises a data-compression-encoded ordinally-arranged set of a plurality of separate video signals that respectively define substantially different sub-spectra of the spatial frequency spectrum of a differential image corresponding to at least a pair of successive frames of original television images; said receiver comprising: a data-compression decoder for separately decoding each of said separate video signals of said data-compression-encoded set to derive a decoded ordinally-arranged set of said plurality of separate video signals; and an inverse differential pulse code modulator (DPCM) for synthesizing a single output video signal defining successive frames of television images corresponding to said original television images in response to said decoded ordinally-arranged set of said plurality of separate video signals being applied as an input thereto.

Claim 22

The receiver defined in claim 18, wherein said differential image corresponds to each pair of immediately successive frames of original television images; and wherein said inverse DPCM includes: a pyramid synthesizer responsive to said decoded set of said plurality of separate video signal being applied thereto for deriving therefrom a first single synthesized video signal corresponding to said differential image; summation means having a first input to which said first single synthesized video signal is applied, a second input to which a feedback video signal is applied, and an output from which a second single synthesized video signal is derived which constitutes the output video signal from said inverse DPCM; and frame delay means to which said first single synthesized video signal is applied as an input, said frame delay means deriving said feedback video signal as an output therefrom, wherein the time delay between corresponding pixels of said first single synthesized video signal and said feedback video signal is one frame period.

Claim 23

The receiver defined in claim 19, wherein: said pyramid synthesizer is a Burt Pyramid synthesizer.