Stmicroelectronics S.r.l.

Stmicroelectronics International N.v.

Daniele Alfonso

Danilo Pietro Pau

Fabrizio Simone Rovati

Emiliano Mario Piccinelli

Tea Anselmo

Alessandro Cremonesi

Daniele Bagni

Elio Privitera

Angelo Schiavone

Fabrizio Basso

Massimiliano Barone

Luca Fanucci

Sumit Johar

Pierluigi Gardella

Tiziana Teresa Signorelli

Luca Celetto

Sergio Saponara

Alexandro Sentinelli

Ravin Sachdeva

Daniele Lavigna

Nicola Capovilla

Amit Kumar

Diego Andres Angulo Pabon

Andrea Graziani

Andrea Cenciotti

A motion analysis is applied to a signal to identify high-motion regions and low-motion regions. An analysis of low-level characteristics of the signal, such as a skin tone or contrast analysis, is used to generate a map of regions with low-level characteristics. The map is filtered to identify regions having both low-level characteristics and low motion. Based on a comparison of the area of the regions having both low-level characteristics and low motion to a threshold, regions of interest are identified. If the area is greater than the threshold, the regions of interest are identified as a function of the high-motion regions and of the regions that have both low-level characteristics and low motion. Otherwise, the regions of interest are identified as a function of the high-motion regions and of regions identified using face detection.

Method and device for video-signal processing, transmitter, corresponding computer program product

An embodiment for encoding digital images includes: partitioning the images into image blocks, subjecting the image blocks to transform into the frequency domain, and, possibly after thresholding resulting in lossy encoding, subjecting the image blocks transformed into the frequency domain to variable length coding to produce compressed encoded image blocks. Transform into the frequency domain may be, e.g., via wavelet transform, such as Haar wavelet transform, and variable length coding may be via Exponential-Golomb codes. An embodiment may also be adapted for transferring picture data over a bus in a system such as, e.g., a System-on-Chip (SoC) by generating compressed encoded image blocks for transfer over the bus and decoding compressed encoded image blocks transferred over the bus.

Method and apparatus for image encoding and/or decoding and related computer program products

The interfacing of coded media data packets in the transfer from and to a peer-to-peer network envisages that the data packets are included in groups of packets that may have lengths different from one another. Each group of data packets is encapsulated in a group of chunks of given length. The media data is made available in the network as a multiplicity of different descriptions of a multiple-description coding of one and the same media content. The terminals of the peer-to-peer network are configured for accessing selectively the multiplicity of different descriptions of said media contents, combining them with one another, or else accessing one of the different descriptions in substitution for another preserving access to the media contents. The data may be made available as base layer and at least one enhancement layer of one and the same media content.

Method and system for distribution of information contents and corresponding computer program product

An embodiment relates to a decoder for decoding CABAC encoded video data in real time for HDTV applications. The decoder comprises a binary arithmetic decoder block for converting an input bit stream into a bin string, a context memory for storing a plurality of context values, and a plurality of finite state machines. Each of the finite state machines is adapted for decoding a particular one of the H.264 syntax elements by providing the binary arithmetic decoder block with an index of the relevant context value within the context memory and by converting the resulting bin stream into a value of the current syntax element. In this manner, a performance of one bin per cycle may be achieved.

System for entropy decoding of H.264 video for real time HDTV applications

An encoder allows generating, starting from a sequence of digital video pictures, a time-scalable encoded bitstream obtained by applying to the pictures, a hierarchical prediction wherein the pictures are organized in Groups Of Pictures (GOPs). The GOPs may include base time layer pictures or Key Pictures for encoding as Inter or Intra, with and without motion-compensated prediction respectively. The GOPs may also include higher time layer pictures adapted to be selectively eliminated to effect time scalability of the encoded scalable bitstream. The encoder may detect scene changes in the sequence of digital video pictures, and, in the presence of a scene change, a first Key Picture after the scene change may be encoded as Intra.

Video coding method and device, related scalable bitstream and computer program product

In an embodiment, a method for converting an input video sequence, comprising digital images organized in frames and operating at a variable frame-rate, into an output video sequence, operating at a pre-set constant frame-rate, storing the input video sequence temporarily and controlling fetching of images of said temporarily stored input video sequence. The method further includes controlling fetching of images from the temporarily stored input video sequence by adjusting an emptying rate to form an intermediate video sequence; and carrying out an operation of motion-compensated interpolation on the intermediate video sequence to form the output video sequence operating at a pre-set constant frame-rate, the emptying rate being adjusted as a function of a number of images of the input video sequence with variable frame-rate temporarily stored. An application is in the reproduction of video sequences in mobile terminals, with the capacity of audio-video conversation, for example for videoconferences.

Method for the frame-rate conversion of a video sequence of digital images, related apparatus and computer program product

Digital video image sequences including slices of macroblocks are encoded by adopting a direct prediction mode, motion-compensated on the basis of motion vectors, chosen from between a direct spatial prediction in which the motion vectors of a given macroblock are obtained from the motion vectors of the macroblocks already encoded within one and the same image, and a direct temporal prediction, in which the motion vectors of a given macroblock are obtained from the motion vectors of the macroblocks belonging to a previously encoded image.

Method and device for digital video encoding, corresponding signal and computer-program product

A method may denoise a digital video signal produced by a photoelectric sensor as a matrix of pixel signals affected by both thermal noise and impulsive noise. The method may include estimating the noise level associated to the pixel signals, and filtering the pixel signals with an attenuation factor that is a function of the estimated noise level.

Method and device for de-noising a digital video signal, and corresponding computer program product

A video sequence including images in the form of digital samples is encoded by: subsampling the video sequence to produce: a number N of multiple descriptions of the video sequence, each multiple description including 1/N samples of the video sequence, and a subsampled version of the sequence, the subsampled version having a resolution lower or equal to the resolution of the N multiple descriptions. The N multiple descriptions and the subsampled version are subjected to scalable video coding (SVC) to produce an SVC encoded signal having a base layer and N enhancement layers predicted from said base layer. The subsampled version of the sequence and the N multiple descriptions of the video sequence constitute the base layer and the enhancement layers, respectively, of the SVC encoded signal.

Encoding and decoding methods and apparatus, signal and computer program product therefor

In an embodiment of a method for converting an input video sequence, comprising digital images organized in frames and operating at a variable frame-rate, into an output video sequence, operating at a pre-set constant frame-rate, it is envisaged to store the input video sequence temporarily and to control fetching of images of said temporarily stored input video sequence. The method moreover envisages: 

A spatial-type de-interlacing process to be applied to a digital image for obtaining a spatial reconstruction. Furthermore, to the digital image there are also applied one or more temporal-type motion compensation de-interlacing processes for obtaining one or more temporal reconstructions, and the spatial reconstruction and the one or more temporal reconstructions are sent to a decision module. The decision module applies a cost function to the spatial reconstruction and the temporal reconstructions and chooses from among the spatial reconstruction and the temporal reconstructions the one that minimizes the cost function. Preferential application is to display systems, in particular displays of a cathode-ray type, liquid-crystal type, and plasma type which use a mechanism of progressive scan.

Method and system for de-interlacing digital images, and computer program product therefor

To carry out de-interlacing of digital images there is provided a spatial-type de-interlacing process to be applied to a digital image for obtaining a spatial reconstruction. Furthermore, to the digital image there are also applied one or more temporal-type de-interlacing processes for obtaining one or more temporal reconstructions, and the spatial reconstruction and the one or more temporal reconstructions are sent to a decision module. The decision module applies a cost function to the spatial reconstruction and the temporal reconstructions and chooses from among the spatial reconstruction and the temporal reconstructions the one that minimizes the cost function. Preferential application is to display systems, in particular displays of a cathode-ray type, liquid-crystal type, and plasma type which use a mechanism of progressive scan.

A system for generating motion vectors in a motion estimator is configured for co-operating with an engine for calculating estimation error for generating motion vectors, according to estimation errors and/or motion vectors previously generated. The system comprises a program memory that contains program data for a motion-estimation algorithm, and a motion-vector memory that contains data identifying said motion vectors previously calculated. The system further comprises an arithmetic and logic unit co-operating with the program memory and motion-vector memory to generate motion vectors in a programmable way to get them to correspond, for example, to predictors deriving from motion-estimation operations already performed on the macroblocks of the reference frame or on the previous macroblocks of the current frame, as updates calculated using the co-ordinates of the motion vectors already issued for the current macroblock, or as absolute motion vectors, generated as such, taking the components directly from the program code.

Programmable system for motion vector generation

A process and a system is described for generating an MPEG output bitstream starting from an MPEG input bitstream. The output bitstream has a resolution modified with respect to the resolution of the input bitstream. In the input bitstream, first portions that substantially do not affect and second portions that do affect resolution variation are distinguished. The second portions are then subjected to a function of modification of the resolution obtained by filtering the second portions in a domain of the discrete cosine transform, and then are transferred to the output bitstream. A corresponding computer program product is also provided.

Process for changing the resolution of MPEG bitstreams, and a system and a computer program product therefor

A method for texture compressing images having a plurality of color components (R, G, B), includes decomposing the images in sub-blocks each including only one color component. At least one first predictor is defined for each sub-block and a respective set of prediction differences is computed for each sub-block. Then the prediction differences for each sub-block are sorted, and a look-up prediction differences palette is set up by defining a look-up prediction error palette. A predetermined code is associated with each column of the error palette.

Growing community of inventors

Magenta, Italy

Daniele Alfonso