Abstract: This relates to a filtering method for processing an image sequence, including for each image parallel binarizations of the gray levels, provided to define a tree structure made of a father node (c.sub.0.sup.1) corresponding to a background (A0) and of successive child nodes (C.sub.h.sup.k) corresponding to associated connected components (B1, D1, F1, C2, E2, G2) and a decision on the elimination or the preservation of each node (C.sub.h.sup.k), followed by an assignment of a gray level value to each picture element. This type of filtering offers the advantage of simplifying the images as well as perfectly preserving the contours, and the decision is based on various proposed simplification criteria, among which the motion-oriented one is particularly attractive.

Abstract: To achieve an enhanced data reduction, a method of fractal image coding, in which method an image to be encoded is divided into range blocks which do not overlap and jointly cover the image contents completely, in which domain blocks comprising areas of the image contents in their original form or in a transformed form are generated from the image contents, and all areas of the image contents are covered by at least one domain block, for each range block the domain block most similar to the range block being searched, and the addresses of the domain blocks found as the most similar and possibly the transform functions used for their generation being transmitted instead of the data of the range blocks, is characterized in that, in a sequence of images, the fractal image coding performed exclusively within the images is only performed with basic images comprised in the image sequence at selectable time intervals, in that, for intermediate images provided between the basic images, the same range and domain block

Abstract: The temperature of a heating element (3), which heating element has a temperature dependent resistance, is measured by comparing the current through it with a user adjustable reference value. The AC voltage across a current sensing resistor (7) in series with the heating element (3) is compared in a comparator (9) with a reference AC voltage generated at a tap (6) of a voltage divider (8). The phase of the logic output signal of the comparator (9) indicates a first state in which the heating element (3) is too hot or is not conducting, or a second state in which the heating element (3) is too cold and conducting. A second comparator (16) detects whether or not the heating element (3) is conducting. A logic gate (18) and a latch (17) switch on the current through the heating element (3) in response to timed pulses from a timing circuit (15) and switch off the current through the heating element (3) when the heating element (3) is conducting and too hot.

Abstract: The invention relates to an image sequence coding method in which images are segmented and coded with respect to their contours and textures. The texture coding step is carried out by means of a lattice vector quantization method adapted to region-based coding schemes where regions have arbitrary shapes, for applications allowing to reach very low bit rates while keeping a good image quality.

Abstract: In order to generate a non-disturbed binary output signal from a disturbed input signal, notably if the latter is periodic in the non-disturbed state, the invention proposes a circuit arrangement which includes two circuit branches which respond to different edge directions of the input signal. Each branch includes two series-connected flipflops and there is also provided a timing member which is common to the two circuit branches. As a result of a coupling of the two circuit branches to one another it is achieved that, after the triggering of one circuit branch, the other branch is blocked. Triggering of the other circuit branch is possible only after the delay time of the timing member has elapsed. It is thus achieved that the edges of the binary output signal have a minimum temporal spacing or that the binary output signal does not exceed a maximum frequency. Furthermore, a transition of the input signal is evaluated also if it occurs prior to the elapsing of the delay time.

Abstract: A variable length decoder (VLD) for processing an input bit stream (e.g., MPEG video) which includes qualifying and non-qualifying types of variable length code words. The VLD includes an input circuit for receiving the input bit stream and providing a sequence of available input bits, a shifter circuit for providing a decoding window that includes one or more code words contained in the sequence of available input bits, a code word length decoding circuit for determining whether or not the decoding window contains a pair of qualifying code words, and for determining the combined length of the pair of qualifying code words and producing a combined length signal representative of the determined combined length, if the decoding window contains a pair of qualifying code words, and further, for determining the length of a leading code word contained in the decoding window and producing a leading word length signal representative of the determined length of the leading code word.

Abstract: There are frames of three types, namely intraframes (I0), predicted frames (P3, P6) and interpolated or bidirectional frames (B1, B2). Each bidirectional frame is decoded twice in succession, a first time (1.sup.1) for obtaining a first field (1e) and displaying it as its decoding proceeds, while the second field is obtained in the course of the same decoding operation but is not used, and subsequently a second time (1.sub.2) for obtaining the second field (1o) and displaying it progressively. As the bidirectional frames are decoded and displayed in lines of blocks, the lines of blocks which are being decoded are located progressively in a memory, each time in place of a line of blocks whose contents have been displayed.

Abstract: A method for providing a secure interface between a decryption engine and a system decoder of a digital receiver, e.g., an MPEG-2 digital television receiver. The system decoder receives an encrypted bitstream and produces a cipher text bitstream which is supplied to the decryption engine via a first parallel data bus which includes a plurality N of parallel bit lines corresponding to respective N bits of the cipher text bitstream. The decryption engine decrypts the cipher text bitstream and produces a plain text bitstream which is supplied to the system decoder via a second parallel data bus which includes a plurality N of parallel bit lines corresponding to respective N bits of the plain text bitstream.

Abstract: A method and apparatus for improved motion vector selection in the motion estimator stage of a video encoder such as an MPEG coder are provided. From a received (re-ordered) image frame sequence, a first stage (14) of the estimator generates a set of candidate motion vectors, suitably from a list of half-pixel candidates from about each of N candidate integer pixels. Those candidate vectors, together with an indication of their respective accuracies, are passed to a vector selection stage (16). A vector bit counter (18) determines the number of bits required per vector and the selection stage (16) then selects those of the candidate vectors for a macroblock line (or "slice") which minimise a cumulative cost function for the line and provide the best trade-off between motion vector accuracy and required bit rate for each block of the line. The derivation of the cumulative cost function and its application to coding mode selection is described.

Abstract: A transmitter for transmitting at least a first and a second main signal component is disclosed, having at least two input terminals for receiving the at least two input signals. The transmitter comprises multiplexer means for multiplexing the at least two input signals to a first input of matrixing means, and to an input of first bitrate reducing means. The first bitrate reducing means is for carrying out a data reduction step on the signal applied to its input and to supply a data reduced version to an output. Expansion means are present, having an input coupled to the first bitrate reducing means, for carrying out a data expansion on the data information applied to its input, so as to obtain a replica of the signal applied to the input of the bitrate reducing means at an output. Said output is coupled to a second input of the matrixing means. An output of the matrixing means is coupled to signal combination means via second bitrate reducing means.

Abstract: Due to a reduction of block artefacts in the reconstructed signal, lapped signal transforms are more suitable than non-lapped transforms for compressing picture and sound signals. Nevertheless, the known lapped orthogonal transform is still accompanied by the artefacts. A system and method are described for transmitting picture or sound signals in which the artefacts remain absent also upon substantial compression and in which orthogonality and phase linearity of the transform is maintained. The transform matrix may be constructed from any arbitrary orthogonal K*K basis matrix C. It is indicated how the basis matrix can be obtained, for example from the DCT.

Abstract: A method for pair-match Huffman transcoding a Huffman-encoded bit stream (e.g., an MPEG digital video bit stream) which includes a plurality of qualifying and non-qualifying types of variable length original code words to be processed by a variable length decoder (VLD) with two-word bit stream segmentation, wherein the VLD has a decoding, window N bits wide. The method includes the steps of pair-match Huffman transcoding at least selected ones of the qualifying types of code words to produce corresponding transcoded code words in such a manner that the combined length of any pair of the transcoded code words is .ltoreq.N. A device for implementing this method, as well as a high-performance variable length decoder with two-word bit stream segmentation which utilizes this method, are also disclosed.

Abstract: System for transmitting n video programs simultaneously through a transmission channel having a predetermined bitrate, comprising in cascade a decoding assembly and an encoding assembly. The decoding assembly consists of n parallel decoding means provided in parallel for decoding the coded input signals corresponding to said programs, and the encoding assembly consists of n encoding means being in cascade with each of said decoding means and each comprising a quantizer having a variable step size and a variable length encoder. The system also comprises means for controlling the step sizes of said encoding means in dependence upon their respective output bitrates, and these control means are also provided for controlling the output bitrates of said encoding means in dependence upon complexity values associated to the coded input signals of each decoding means with respect to the joint complexity of the plurality of said coded signals.

Abstract: A method of transcoding coded video signals and a corresponding device for implementing this method are described. The device comprises decoding and encoding sub-assemblies (201, 202) and, within the encoding sub-assembly, quantizer (23) and bitrate controller. This bitrate controller comprise, before the quantizer (23), circuit for computing a local target value for encoding each sub-picture of the images to which video signals correspond and circuit for controlling the step size during the quantizing step.

Abstract: A method and apparatus for dynamically allocating the available bandwidth of a common transmission channel of a multiplexed system among multiple encoders in such a manner as to maximize and equalize the quality of the encoded data output by all of the encoders, while also preventing underflow or overflow of encoder or decoder buffers at each end of the common transmission channel, and moreover, while also ensuring compliance with (i.e., without violating) the data encoding and transmission protocol utilized by the system. Further, the bandwidth of the common transmission channel is preferably allocated using an algorithm that does not impose any constraints on the size of the encoder or decoder buffers, other than any constraints specified by the data transmission protocol employed in transmitting the encoded data over the common transmission channel.

Abstract: A method is provided for encoding of digital video signal segments which are interleaved to form a single system stream (video STD plus audio STD). In order to enable seamless joining of the segments, control of the video system layer buffer loading is related to the loading of the video layer buffer (VBV occupancy). Starting from an idealised match at the point the first picture is to be read out of the decoder buffer, the point at which the first picture is to be read out of the system layer buffer is adjusted by an amount determined by the predicted duration of the segment and the point at which the last frame of the segment is due to be read out. The adjustment prevents possible overlap of segment timing values derived initially from the idealised match case. The technique has particular application to interactive multimedia applications where real time splicing of video clips is required with minimal visual discontinuity.

Abstract: An ashtray houses a fan (20) which drives air through two filters (30, 36). Air is drawn in from an ashtray well (6) and expelled from openings (17) in the base of the ashtray. A recirculation path (A.sub.1, A.sub.2, A.sub.3) is provided for pre-filtering by the first filter 30.

Abstract: A device and method for removing co-channel conventional television signal interference from a digital signal by using an-adaptive filter which tunes to the periodic components of the conventional television signal. The output of the adaptive filter is then subtracted from the received signal to produce an error signal. This error signal is then used to recursively update the taps of the adaptive filter. The invention also relates to the use of the adaptive filter in conjunction with a DFE.

Abstract: Device for segmenting data corresponding, for example to image sites, for transmitting and/or storing these images at a low rate. This device comprises a sub-assembly for initial segmentation into classes, supplying an initial set of labels. In two parallel paths (201, 202), a sub-assembly (200) determines conditional probabilities, notably on the basis of a propagation factor L whose parameters are defined by specific relations. These probabilities are used in a labelling sub-assembly (300) which supplies a new set of labels. Finally, a sequencing sub-assembly (400) determines the replacement of the initial set of labels at the input of the sub-assembly (200) by the new set and continues or, in contrast, discontinues this iterative process of determining new labels.

Abstract: The picture quality of an MPEG-encoded video signal is improved considerably by reducing the coefficients after quantization in such a way that the "Lagrangian cost" D+.lambda.R (D is distortion, R is bitrate) is minimal for a given value of the "Lagrange multiplier" .lambda.. The distortion decreases because the step size is reduced while the bitrate remains equal.