Kokusai Denshin Denwa Co., Ltd.

Kokusai Denshin Denwa Kabushiki Kaisha

Fujitsu Corporation

Nec Corporation

Nippon Telegraph and Telephone Corporation

British Telecommunications Public Limited Company

Denka Company Limited

Masahide Kaneko

Yoshinori Hatori

Atsushi Koike

Kiichi Matsuda

Mutsumi Ohta

Naoki Mukawa

Yoichi Kato

Seiichi Yamamoto

Toshio Koga

Mark A Shackleton

Yoshitaka Taniguchi

Ryozo Nonogaki

Toshitaka Yamagata

William John Welsh

Norio Higuchi

Kosuke Wada

Tsutomu Miyasato

A picture synthesizing apparatus, and method for synthesizing a moving picture of a person's face having mouth-shape variations from a train of input characters, wherein the method steps comprise developing from the train of input character a train of phonemes, utilizing a speech synthesis technique outputting, for each phoneme, a corresponding vocal sound feature including articulation mode and its duration of each corresponding phoneme of the train of phonemes. Determining for each phoneme a mouth-shape feature corresponding to each phoneme on the basis of the corresponding vocal sound feature, the mouth-shape feature including the degree of opening of the mouth, the degree of roundness of the lips, the height of the lower jaw in a raised and a lowered position, and the degree to which the tongue is seen. Determining values of mouth-shape parameters, for each phoneme, for representing a concrete mouth-shape on the basis of the mouth-shape feature; and controlling the values of the mouth-shape parameters for each phoneme, for each frame of the moving picture in accordance with the duration of each phoneme, thereby synthesizing the moving picture having mouth-shape variations matched with a speech output audible in case of reading the train of input characters.

Mouth shape synthesizing

A moving image with a plurality of continuous frames is processed by using a plurality of image processing units (5-1 through 5-n), which operate simultaneously. A frame identification designator (4) attaches an identification number cyclically to each frame of the input image signal. A frame identification selector (6) in each image processing unit inputs an image frame selectively based upon the identification number, and transfers the selected frame to a corresponding image processing module. The image processing units (5-1 through 5-n) typically carry out either the same processing on respective image frames, or a plurality of processing to a single frame. An additional processor network (8) may be coupled with the image processing units for high-level processing including recognition and/or understanding of an image.

Image analysis system

Method and device for frame interpolation of a moving image

A system for detecting three-dimensional motion of a three-dimensional object by obtaining, for each block, a velocity vector from two temporally successive images of a moving picture including the three-dimensional object. A three-dimensional model is prepared in correspondence to the solid configuration of the three-dimensional object, an area is obtained in the previous frame picture where the three-dimensional object and the three-dimensional model overlap, the area is divided into a plurality of blocks, a velocity vector between the current input image and the previous frame image is obtained for each block, depth information in the block is obtained from the three-dimensional model, and three-dimensional motion parameters of the three-dimensional object, that is, its rotational movement and translational movement components, are obtained from the velocity vectors and depth information thus obtained in the plurality of blocks, whereby the three-dimensional motion of the three-dimensional object is detected.

Method and apparatus for detecting motion of moving picture

An intelligent coding system for a picture signal is disclosed, in which a feature picture representing features of a picture is produced on the basis of an input picture signal, information of the position and the shape of each element forming the picture is extracted from the feature picture and then the information thus extracted is organized concisely by a small number of parameters. According to another aspect of the present invention, consistency (or continuity) of variations in the position and the shape of each element can be held throughout a continuation of frames of motion picture signals.

Intelligent coding system for picture signal

A moving image is coded through a waveform coding predictor (2), and a model-based coding predictor (3) which is mainly applicable to a head portion of a person, the coded outputs of said two predictors are compared (5) with an input current image for each block element which has for instance 8.times.8 or 16.times.16 pixels. A block element coded for the smaller difference from an input current image is taken, and is transmitted to a receiving side through differential coding (6,7). As the information compression ratio is high, a moving image can be transmitted through conventional telephone line, or a narrow band circuit.

Hybrid coding system for moving image

In an encoding system responsive to a sequence of coefficient signals which results from a predetermined linear transform and which is divided into a sequence of blocks, anterior processing is carried out prior to quantization about the coefficient signals in each block to determine a significant area of the coefficient signals by comparing each level of the coefficient signals with a threshold level. Only the coefficient signals within the significant area are judged to be valid and produced as a sequence of significant coefficient signals from a classifying circuit (20) to be quantized by a quantizer (15) into a sequence of quantized signals. The quantized signal sequence is subjected to posterior processing to produce a sequence of encoded signals. The significant area may be decided by the use of a selected one of zone detection, zigzag scanning, and adaptive scanning. In the posterior processing, the quantized signals may be encoded into the encoded signals by the use of a plurality of code sets or tables.

Encoding system capable of accomplishing a high efficiency by anterior

A motion compensation difference interframe or intra-frame coding system includes a block data redundancy compression and coding unit, a PIXEL data coding unit and a prediction data generator unit. The block data redundancy compression and coding unit codes a motion compensated (MC) difference between an input image in a block and a motion predicted image from the prediction data generator unit, and transmits coded data to a receiver. The PIXEL data coding unit receives an error between the MC difference and a decoded MC difference from the block data redundancy compression and coding unit, rearranges the error in PIXEL data and codes the PIXEL error, when the error is greater than a predetermined value. The PIXEL coded data is also transmitted to the receiver. The prediction data generator unit generates predicted block data of the motion of the image. The redundancy compression and coding unit may include a filter circuit rejecting pulse components contained in the MC difference.

System for coding and transmitting motion image signals

In an encoder responsive to a sequence of coefficient blocks each of which is composed of a plurality of orthogonal transform coefficients subjected to predictive encoding and orthogonal transform, the coefficient blocks are classified by a classification circuit (55) into a plurality of classes to produce a control signal (CONT) representative of the classes. A quantization unit (60), a code conversion unit (61), and a coefficient selection unit (62, 64) are controlled by the control signal to select quantization characteristics, code conversion characteristics, and coefficient selection characteristics. The quantization unit comprises a plurality of quantizers corresponding to the quantization characteristics while the code conversion unit and the coefficient selection unit comprise a plurality of code converters and a plurality of selection tables corresponding to the code conversion characteristics and the coefficient selection characteristics, respectively. Consequently, the coefficients are adaptively encoded into a reduced amount of encoded codes on encoding a moving image.

Encoder capable of faithfully and adaptively encoding a moving image

A motion vector V.sub.x of a picture block X between a current input picture and a preceding frame picture is corrected to V.sub.p, when said motion vector V.sub.x differs from a motion vector V.sub.r of adjacent blocks of said current input block, and (S.sub.v.sbsb.x &lt;T.sub.hl, and S.sub.v.sbsb.p .gtoreq.T.sub.h2) is not satisfied, where S.sub.v.sbsb.x is a block difference of the block X between the current frame F.sub.0 and the preceding frame f.sub.2 for said motion vector V.sub.x, V.sub.p is one of motion vectors V.sub.r giving the minimum value of S.sub.v.sbsb.r, and S.sub.v.sbsb.p is a block difference for said v.sub.p. The present invention provides correct motion vector of a moving object on a screen irrespective of noise in a picture signal and sudden intensity change. The present invention is applicable to a frame interpolation in a television standard system conversion system, and an efficient coding of a television signal.

Motion vector detection system of a moving object on a screen

A coded picture signal re-quantization system, in which a parameter such as a variance or a mean value of picture element values in a region to be re-quantized is calculated on the basis of sub-sampled picture element values and, according to this parameter, a quantizer is adaptively varied for each region and, further, quantization of thinned-out picture elements is performed or stopped as required, thereby effecting conditional re-quantization.

System for re-quantization of coded picture signals

A system for detecting the vector of motion of a moving object on a picture screen, in which there is prepared, in addition to a picture memory for storing picture signals, a background memory for storing background picture signals. From the background memory, picture elements of a block #0b which lies at the same position on the screen as the current input block are read out and an inter-block difference B(#0', #0b) between the current input block #0' and the block #0b is calculated. If this gives rise to an uncovered background, then the above inter-block difference B(#0', #0b) between the picture memory and the background memory becomes smaller than any other inter-block differences (that is, the degree of similarity becomes the highest), and consequently, it is possible to judge that a new background portion has appeared. Incidentally, in connection with the background portion except the uncovered background, for example, a portion A' data in the picture memory and data in the background memory become identical, excepting noise components. Accordingly, in connection with this portion, the contents of the block #0 in the picture memory and the block #0b in the background memory differ little from each other, so that even if either the inter-block difference B(#0', #0) or B(#0', #0b) is chosen as a result of comparison of their magnitudes, substantially no characteristic deterioration is caused. For instance, in case of using the result of detection of movement for the movement-compensated predictive coding operation, the coding efficiency is hardly affected, even if either the block #0 or #0b is used for prediction.

Growing community of inventors

Tokyo, Japan

Masahide Kaneko