Stmicroelectronics S.a.

Stmicroelectronics (grenoble 2) Sas

Stmicroelectronics (rousset) Sas

Institut Polytechnique De Grenoble

Université Pierre-mendès France

Marina Nicolas

Fritz Lebowsky

Frédérique Crete

Eric Van Reeth

Jerome Roussel

Pascal Bertolino

Patricia Ladret

A method for compressing a data block including sets of homologous components may include selecting a designated component from the data block, and compressing non-designated components with a measurable loss less than or equal to a threshold. The method may further include compressing the designated component based upon at least a selection of values from among the values of the homologous designated components associated with the data of the block.

Compression and decompression of video data

A source image is transformed into a destination image having a target aspect ratio. A reference region in the source image is defined. An extended region of interest of the source image having the target aspect ratio and containing the reference region is defined. A set of candidate image regions of increasing resolutions from the extended region of interest is determined, each having the target aspect ratio and containing the reference region. Candidate image regions are scaled to form a candidate target images. A quality metric is used to select a target image providing the best quality metric value.

Method for combined transformation of the scale and aspect ratio of a picture

An image processing method, implemented in a calculator, includes applying a process by group of pixels to an original image. For each group of pixels, calculating a cumulative sum of value or position differences of the pixels of the group of pixels, and for each group of pixels, allocating in a final image signal a pixel position of the group of pixels to each pixel value of the group of pixels so as to minimize the cumulative sum of differences calculated for the group of pixels according to the differences calculated. For each group of pixels, determining a filtering intensity according to the cumulative sum of differences calculated for the group of pixels, and applying to the group of pixels a filtering having the filtering intensity.

Method for selecting a pixel positioning solution to obtain an optimal visual rendition of an image

A method involving: detecting in an image a first type of artifact (A) to generate a first artifact value (G) and a first confidence level (C); detecting in the image a second type of artifact (A) to generate a second artifact value (G) and a second confidence level (C); and performing correction of the first type of artifact (A) in the image based on the first and second artifact values (G, G) and the relative values of the first and second confidence levels (C, C).

Detection and correction of artefacts in images or video

There is provided a method of measuring the similarity of parts of digital image files (IF, IF,IF-)), which comprising the steps of calculating a first change value in a similarity between pixel values in a first segment (BIF) of a first digital image file (IF) and in a second segment (BIF) of this first file (IF), said first and second segments of said first file being spatially separated by a first translation vector (td), calculating a second change value (v) in the similarity between pixel values in a first segment (BIF) of the second digital image file (IF, IF-) and in a second segment (BIF) of the second file (IF, IF-), said first and second segment of said second file being spatially separated a second translation vector (td), wherein the first segment of the first file corresponds to substantially same part of the image as the first segment of the second file and wherein the second segment of the first file corresponds to substantially the same parts of the image as the second segment of the second file, and calculating a structure evolution value indicative of the similarity between the first and second changes.

Fidelity measurement of digital images

A motion estimation method and device are provided for processing images to be inserted, between a preceding original image and a following original image, into a sequence of images. Each image is divided into pixel blocks associated with motion vectors. For a current block of an image being processed, motion vectors associated with blocks of the image being processed and/or associated with blocks of a processed image are selected. Candidate vectors are generated from selected motion vectors. An error is calculated for each candidate vector. A penalty is determined for a subset of candidate vectors on the basis of the values of the pixels of the pixel block in the preceding original image from which the candidate motion vector points to the current block and/or on the basis of the values of the pixels of the pixel block in the following original image to which the candidate motion vector points from the current block.

Method and device for image interpolation systems based on motion estimation and compensation

A motion estimation method and device are provided for processing images to be inserted, between a preceding original image and a following original image, into a sequence of determined images. Each image is divided into pixel blocks and a motion vector is associated with each block of a processed image. Motion vectors associated with respective blocks of the image are processed and motion vectors associated with respective blocks of a preceding processed image (spatial and temporal motion vectors) are selected with respect to a current block of an image being processed. Candidate motion vectors are generated from the motion vectors selected. One vector is elected from among the candidate vectors. Information associating the elected vector with the block is stored. A temporal motion vector is selected only if it satisfies a determined selection criterion based on the orientation of the motion vector.

Method and device for generating candidate motion vectors from selected spatial and temporal motion vectors

Method and devices are provided for processing an image in which a row of pixels with associated light intensity values successively alternates with a row of pixels to which values are to be associated. In each current row of pixels with associated values, pixels are selected which satisfy a selection rule based on the values for the pixels in the current row with associated values and on the values from the previous and/or the next row with associated values, and segments of pixels are detected. In this manner a set of segments is obtained relative to the image being processed, from which a graph of segments is created by linking the segments according to a linking rule. Then the respective values to be associated with at least some of the pixels in the rows to be assigned values are determined by interpolation from segments in the graph which form a path in the graph.

Image deinterlacing

An image comprises columns and rows of blocks of pixels. In each row and/or column, a series of pairs of blocks of pixels comprise first and second adjacent blocks of pixels, with the second block in a previous pair of blocks corresponding to the first block in a next pair of blocks. For at least a part of the pairs of blocks in at least a part of the rows and/or columns of blocks in the image, a method determines an indication of a block boundary between the first and second blocks. The method decides based on the determined indication whether the block boundary is a visible or invisible block boundary. These steps are then repeated for a next pair. If at least one invisible block boundary is present between two visible block boundaries, it is decided that a uniform image distortion zone has been detected.

Image noise detection

Artifacts of an incident digital image including pixels carrying information are reduced by determining, for certain pixels being considered from the image, displaced pixels. A displaced pixel associated with a pixel being considered is situated at a location that is displaced with respect to the location of the pixel being considered. Substitution information is determined by taking into account the variations between each piece of information carried by pixels situated at locations adjacent to the pixel being considered. The pixel being considered is then selectively replaced by a substitution pixel equal to the displaced pixel or to a combination of the displaced pixel and the pixel being considered, depending on the value of the substitution information.

Method and device for reducing blocking artifacts in a compressed digital image without reducing clarity of edges

The invention provides for a process and a device for de-interlacing a video signal, wherein at output (S) is produced a signal (Sde) of video images de-interlaced by interpolating the pixels missing from the interlaced video signal presented at input (E), the interpolation on the output signal (Sde) being composed selectively () from a spatial interpolation (), based on a transition detection and from a temporal interpolation () with a decision being made on the variable degree of presence of spatial interpolation and/or of temporal interpolation in the output signal (Sde), the decision being made as a function of a motion detection in the relevant area of the image, wherein the decision is made additionally as a function of a detection of the detail () in a relevant area of the image.

Growing community of inventors

Voreppe, France

Marina Nicolas