Coding and decoding method and device

A method for encoding video having an image sequence whereby images in the image sequence are encoded in a scaled manner such that resulting encoded video contains data which ensures representation of the images in a plurality of different levels of image resolution defined by a number of pixels per image representation, the method comprising:

• temporarily creating a first block structure for a first resolution level and a second block structure for a second resolution level, whereby the first resolution level is lower than the second resolution level, the first and second block structures being created in a block-based manner such that, starting from one block, the block is subdivided into partial blocks and then sub-blocks to describe a motion possibility in the image sequence, the motion possibility being described by more finely subdividing areas of motion:
• comparing the first block structure with the second block structure to determine structural differences including motion and texture between the first and second block structures;
• creating a modified second block structure based on the structural differences, whereby the modified second block structure constitutes a subset of the second block structure;
• and
• using the modified second block structure to encode the image in the image sequence, wherein
• the sub-blocks have characteristics which are analyzed to determine differences,
• only the sub-blocks of the second block structure whose block size is at least a definable threshold value are transferred into the modified second block structure, and
• the threshold value is defined such that the threshold value represents a ratio between the block size of one sub-block of the second block structure and the block size contained in a region of the first block structure, the threshold value being assigned to the smallest sub-block of the region of the first block structure.

A method for encoding video having an image sequence whereby images in the image sequence are encoded in a scaled manner such that resulting encoded video contains data which ensures representation of the images in a plurality of different levels of image resolution defined by a number of pixels per image representation, the method comprising:

• temporarily creating a first block structure for a first resolution level and a second block structure for a second resolution level, whereby the first resolution level is lower than the second resolution level, the first and second block structures being created in a block-based manner such that, starting from one block, the block is subdivided into partial blocks and then sub-blocks to describe a motion possibility in the image sequence, the motion possibility being described by more finely subdividing areas of motion;
• comparing the first block structure with the second block structure to determine structural differences including motion and texture between the first and second block structures;
• creating a modified second block structure based on the structural differences, whereby the modified second block structure constitutes a subset of the second block structure;
• and
• using the modified second block structure to encode the image in the image sequence,
• wherein the modified second block structure of the second resolution level is used as a first block structure of a third resolution level, the second resolution level having a smaller number of pixels than the third resolution level.

A method for decoding an encoded image sequence, wherein a scalable representation of the image sequence is created based upon the second modified block structures contained in the image sequence that were created according to a method for encoding video whereby images in the image sequence are encoded in a scaled manner such that resulting encoded video contains data which ensures representation of the images in a plurality of different levels of image resolution defined by a number of pixels per image representation, the method comprising:

• temporarily creating a second block structure for a second resolution level, whereby the second resolution level is higher than a first resolution level, the second block structure being created in a block-based manner by a process comprising:
• starting from one block, the block is subdivided into partial blocks and then sub-blocks;
• time-predicting the second block structure to produce a predicted block structure, the second block structure being time-predicted based on at least one other image in the image sequence;
• determining errors in the predicted block structure;
• and
• excluding from the second block structure information not relating to the errors in the predicted block structure;
• comparing the second block structure with a first block structure corresponding to the first resolution level, the first and second block structures being compared to:
• recognize further block structure divisions included the second block structure but not the first block structure;
• and
• determine the sizes of the further block structure divisions;
• creating a modified second block structure by eliminating from the second block structure all further block structure divisions larger than a predefined threshold value whereby the modified second block structure constitutes a subset of the second block structure;
• and
• using the modified second block structure to encode the image in the image sequence.

A method for encoding video having an image sequence whereby images in the image sequence are encoded in a scaled manner such that resulting encoded video contains data which ensures representation of the images in a plurality of different levels of image resolution defined by a number of pixels per image representation, the method comprising:

• temporarily creating a first block structure for a first resolution level and a second block structure for a second resolution level, whereby the first resolution level is lower than the second resolution level, the first and second block structures being created in a block-based manner such that, starting from one block, the block is subdivided into partial blocks and then sub-blocks to describe a motion possibility in the image sequence, the motion possibility being described by more finely subdividing areas of motion;
• comparing the first block structure with the second block structure to determine whether sub-blocks in the second block structure are four times smaller than sub-blocks in the first block structure;
• creating a modified second block structure by supplementing the first block structure to additionally include only the sub-blocks of the second block structure that are four times smaller than sub-blocks of the first block structure;
• and
• using the modified second block structure to encode and transmit the image in the image sequence.

An encoder for creating an encoded image sequence, whereby images in the image sequence are encoded in a scaled manner such that resulting encoded video contains data which ensures representation of the images in a plurality of different levels of image resolution defined by a number of pixels per image representation, the encoder comprising:

• means for temporarily creating a second block structure for a second resolution level, whereby the second resolution level is higher than a first resolution level, the second block structure being created in a block-based manner by a process comprising:
• starting from one block, the block is subdivided into partial blocks and then sub-blocks;
• time-predicting the second block structure to produce a predicted block structure, the second block structure being time-predicted based on at least one other image in the image sequence;
• determining errors in the predicted block structure;
• and
• excluding from the second block structure information not relating to the errors in the predicted block structure;
• means for comparing the second block structure with a first block structure corresponding to the first resolution level, the first and second block structures being compared to:
• recognize further block structure divisions included the second block structure but not the first block;
• and
• determine the sizes of the further block structure divisions;
• means for creating a modified second block structure by eliminating from the second block structure all further block structure divisions larger than a predefined threshold value whereby the modified second block structure constitutes a subset of the second block structure;
• and
• means for using the modified second block structure to encode the image in the image sequence.