Bell Telephone Laboratories

Frank W Mounts

Arun N Netravali

Birendra Prasada

Kenneth A Walsh

Edward George Bowen

The efficiency with which a video signal is encoded is increased by examining successive blocks of M picture elements to determine which K pels have frame differences which do not exceed a predetermined threshold. The values of these K pels are intentionally altered, and the pels are applied to a transform circuit which generates N.ltoreq.M corresponding coefficient values. Only N-K of the coefficient values, and a code indicating the unchanged pel locations are then encoded for transmission. The decoder (and the feedback loop of the encoder) is arranged to recover the original picture using only the frame differences for pels which have changed significantly.

Video interframe transform coding technique

A series of bits which represent the intensity values of elements of a facsimile picture are preprocessed to eliminate runs of length one, and then applied to a predictive encoder which treats values for alternate pels in two distinct ways. Specifically, the value of every odd pel is predicted, based upon the intensity values of some surrounding picture elements. Each prediction is compared with the true intensity value to yield an error signal. The encoder is also arranged to generate a reference signal which indicates whether the confidence in each prediction for odd pels is high or low. The error values for odd pels within a predefined reordering interval are then sorted or reordered into two groups, dependent upon the reference value. On the other hand, the intensity values of even pels are predicted only when the values cannot be precisely determined from the two previous and subsequent values; the error values for even pels are assembled into a third group. When data for an entire reordering interval has been processed, each of the groups is run length coded for transmission. By encoding even pel values only when they are needed, the amount of data required to represent the facsimile picture is reduced, at the expense of only a slight loss in detail, which is not objectionable. Further, since the statistics of the groups formed by reordering are different, different code dictionaries can be used for run length coding.

Run length encoding of facsimile pictures

A facsimile signal encoding technique reduces the amount of information needed to represent a picture by (a) generating an error signal (on line 1120) which indicates whether an intensity value prediction corresponds to the true intensity value, (b) generating a reference signal (on line 1005) indicating whether the confidence in each prediction is high or low, (c) reordering the error signal values (logic 1010) into high and low confidence groups in accordance with the reference signal, and (d) assigning code words (encoder 1016) to represent the run lengths of the reordered data. Advantageously, different code dictionaries are selected (line 1019) for code assignment in each of the groups, each dictionary being tailored to the type of runs expected in each group. Also, code words for certain runs may be dropped, and the lengths of other runs may be combined (by logic 1015) before run length coding, again increasing encoder efficiency. A receiver (FIGS. 16-18) for decoding the run lengths is described.

Run length encoding of facsimile signals

When a video signal is encoded using differential pulse code modulation (DPCM) techniques, the frequency of occurrence of the various quantizer output or representative levels is non-uniform; usually, low levels occur more frequently than high levels, and this characteristic can be used to advantage by applying the DPCM output to a variable length encoder before transmission. The present invention increases this advantage by intentionally forcing the quantizer output to particular levels when doing so is not harmful to the picture fidelity. By so doing, the entropy of the encoded signal is reduced by as much as 25 percent.

Video signal encoding

A facsimile scene typically includes a plurality of lines, each line having a plurality of picture elements (pels). Often, pel signals are adaptable for run-length coding, a run being one or more successive pels having the same brightness level. To increase the length of a run and hence to permit a more efficient use of a transmission link, a processor is disclosed for permuting a measure of the pel signals responsive to a reference signal. Illustratively, the reference signal is the pel signal from a previous line. In one exemplary bi-level pel signal embodiment, the measure is the current pel signal. Specifically, if the reference pel signal is a logic one, the current pel signal is loaded beginning at one end of a memory; if the reference pel signal is a logic zero, the current pel signal is loaded beginning at the other end of the memory. In a second embodiment, the measure is an error signal for indicating a difference between the reference signal and the current pel signal. But for loading the error signal rather than the current pel signal, the operations aforedescribed are again followed.

Processor for increasing the run-length of facsimile data

A facsimile scene typically includes a plurality of lines, each line having a plurality of picture elements (pels). Often, pel signals are adaptable for run-length coding, a run being one or more successive pels having the same brightness level. To increase the length of a run and hence to permit more efficient use of a transmission link between transmitter and receiver, a processor arrangement is disclosed for permuting a measure of the pel signals responsive to a reference signal. The reference signal is a calibration signal. The measure is an error signal for indicating a difference between the current pel signal and a prediction thereof. In an illustrative bi-level facsimile system embodiment, if the calibration signal is a logic one, the error signal is loaded beginning at one end of a memory; if the calibration pel signal is a logic zero, the error signal is loaded beginning at the other end of the memory. The loaded permuted error signal having an increased run length is then sequentially read from one end of the memory for extension to a state of the art run-length coder.

Processor for increasing the run-length of digital signals

In PCM and DPCM systems, a Hadamard orthogonal transformation of picture element signals is known to offer a means for reducing the bit rate for transmission of the signals. The first Hadamard coefficient, called H.sub.1, equals the sum of the picture element signals. Accordingly, still further reduction in the transmission bit rate can be achieved by efficiently coding H.sub.1. The disclosed coder includes a control signal generator for providing a control signal corresponding to a predetermined function of the quantized Hadamard coefficient signals. Several illustrative arrangements are disclosed. In three arrangements, the respective control signals are employed to adaptively select one of a plurality of quantizers, the selected quantizer for coding H.sub.1. In a fourth arrangement, the control signal is employed to predict H.sub.1, the predictor control signal being thereafter used for coding H.sub.1.

Growing community of inventors

Colts Neck, NJ, United States of America

Frank W Mounts