Canon Kabushiki Kaisha

Canon Research Centre France

Hervé Le Floch

Frédéric Maze

Franck Denoual

Jean Le Feuvre

Cyril Concolato

Eric Nassor

Eric Nguyen

Christophe Gisquet

Naël Ouedraogo

Stéphane Baron

Xavier Henocq

Ioana Donescu

Cédric Riou

Thomas Ronan

The invention relates to encapsulating tiled timed media data in a server and providing a timed media data bit-stream from encapsulated tiled timed media in a client device. The tiled timed media data comprises timed samples and each timed sample comprises a plurality of spatial subsamples. After having selected at least one spatial subsample from amongst the plurality of spatial subsamples of one of the timed samples, one tile track comprising the selected spatial subsample and one corresponding spatial subsample of each of the other timed samples is created for each selected spatial subsample. One composite track comprising at least one extractor identifying at least one of the created tile tracks is created. Each created track is independently encapsulated in at least one media segment file.

Method, device, and computer program for encapsulating partitioned timed media data

To generate 3D representation of a scene volume, the present invention combines the 3D skeleton approach and the shape from silhouette approach. The present invention efficiently works on complex scenes like sport events with multiple players in a stadium, with an ability to detect a wide number of interoperating 3D objects like multiple players.

3D representation reconstruction from images using volumic probability data

To generate 3D skeletons of a scene volume, a first approach retrieves probabilities from source images capturing the scene to detect parts of the objects, next merges the probabilities in 3D space. The merged or joint probabilities are then used to robustly detect the parts of the objects in the 3D space. This approach advantageously reduces complexity of the 3D skeleton reconstruction. In particular, at 2D level (i.e. on the source images), it avoids conflict resolutions to be performed between conflicting parts detected from different source images. It turns that real time reconstructions and subsequent displays or human motion detection are better achieved. The present approach efficiently works on complex scenes like sport events with multiple players in a stadium, with an ability to detect a wide number of interoperating 3D objects like multiple players.

3D skeleton reconstruction from images using volumic probability data

A method of processing video data comprising subdividing a video stream into image portions, encoding the image portions with high resolution in respective encoded video streams and reassembling a selection of encoded video streams in order to display a region of interest with high resolution. Embodiments of the invention provide high resolution display of specific region of interest without a priori knowledge of the position of the region of interest.

Method of displaying a region of interest in a video stream

The present invention relates to a method and device for generating a description file about a video sequence at a server device for a client device, to retrieve a video segment and containing a region of interest. The method comprising for each video segment, determining a time interval during which the detected region of interest is spatially included in a same frame region in the video sequence; and generating a description file comprising spatial information describing the frame region and temporal information describing a duration at least equal to the determined time interval.

Method and device for generating a description file, and corresponding streaming method

A method for streaming data corresponding to a video, each of a plurality of frames of the video being divided into n spatial tiles, the method includes transmitting, a description file to a client device, the description file including data about a spatial position of each of the n spatial tiles and including data about at least n URLs corresponding to the n spatial tiles, determining by the client device one or more URLs according to the data about the spatial position included in the description file, receiving from the client device, by a server, one or more request messages for requesting one or more spatial tiles, each request message comprising one of the URLs determined by the client device, and transmitting to the client device, by the server, video data corresponding to the one or more spatial tiles, in response to one or more the request messages.

Method and corresponding device for streaming video data

At least one embodiment of a method of encoding a plenoptic image includes: decomposing the plenoptic image into sub-aperture images, encoding at least a part of a first sub-aperture image using a set of encoding parameters, and encoding at least a part of a second sub-aperture image using a restricted number of parameters among the set of encoding parameters. Embodiments make it possible to reduce the encoding complexity and run-time while keeping good encoding performances.

Optimized plenoptic image encoding

An encoder extracts motion vectors from a frame I(t&#x2212;1) preceding the frame I(t) being encoded and processes them to create an estimated motion vector field I(t) for the frame being encoded. A minimized difference between the motion vector field of the frame being encoded and the estimated motion vector field is used to generate transform parameters, which are transmitted to the decoder as auxiliary information along with the usual motion prediction information. The decoder receives the transform parameters. The decoder also creates an estimated motion vector field I(t) for based on a preceding frame I(t&#x2212;1) and applies the transform parameters to the estimated motion vector field to obtain missing motion vectors. The motion vector field rebuilt using the reconstructed missing motion vectors is used for subsequent error concealment/decoding/displaying.

Method and device for error concealment in motion estimation of video data

A method for compressing a set of images comprising a plenoptic image and at least one refocused image obtained from the plenoptic image for a given focus, by generating a set of data comprising a compressed version of the plenoptic image, some focusing parameter to build a refocused image from this plenoptic image and a compressed residual data to restore the refocused image in its plain quality. Accordingly, the set of images can be compressed in an efficient way, preserving the quality of the shoot using the focus chosen by the photographer.

Method and apparatus for compressing or decompressing light field images

In a method of encoding a video frame to be transmitted to a plurality of clients over a plurality of network connections, a set of parameters is associated with each connection. The method comprises: This method advantageously provides an easy and straightforward way for selecting the appropriate parameters that need to be taken into account for controlling the rate of the generated bitstream, particularly when multiple parameters are associated with each connection.

Growing community of inventors

Rennes, France

Hervé Le Floch