Media File Tile Indexing for Fast Partial Decoding

Solution Overview

Problem

Current HEVC media file formats require decoding the entire picture area to determine the position of a specific tile, leading to increased decoding time and power consumption when only a partial area is extracted and played back.

Innovation Solution

Incorporating a slice index box, tile index box, or tile offset box in the media file format to store the ordinal position or offset of each tile, allowing direct access to coded data without decoding unnecessary slices, thereby reducing decoding time and power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the entire picture area is decoded to determine the position of a specific tile, then the position information can be obtained, but the decoding time and power consumption increase

Engineering Contradiction:

Improveposition informationVSAvoiddecoding time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing tile position information in the media file metadata during the coding phase. The tile offset values are computed and saved in advance, so that during playback, the decoder can directly retrieve the position information without performing full picture decoding. This resolves the contradiction by obtaining position information (improving measurement precision) while avoiding the time cost of decoding the entire picture (reducing loss of time).

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent segments the picture into multiple tiles and stores offset information for each tile independently in the metadata. This segmentation allows the system to access only the specific tile needed rather than decoding the entire picture. By dividing the picture into independent tile units with stored position information, the system achieves fast position determination without full decoding, thus resolving the contradiction between obtaining position information and minimizing decoding time.

Inventive Principle:

Principle #1Segmentation

2Measurement precision

If the entire picture area is decoded to determine the position of a specific tile, then the position information can be obtained, but the power consumption increases

Engineering Contradiction:

Improveposition informationVSAvoidpower consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing tile position information in the media file metadata during the coding phase. The tile offset values are computed and saved in advance, so that during playback, the decoder can directly retrieve the position information without performing full picture decoding. This resolves the contradiction by obtaining position information (improving measurement precision) while avoiding the power consumption of decoding the entire picture (reducing use of energy).

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent segments the picture into multiple tiles and stores offset information for each tile independently in the metadata. This segmentation allows the system to access only the specific tile needed rather than decoding the entire picture. By dividing the picture into independent tile units with stored position information, the system achieves fast position determination without full decoding, thus resolving the contradiction between obtaining position information and minimizing power consumption.

Inventive Principle:

Principle #1Segmentation

3Ease of operation

If slices are not divided into tiles, then the media file structure is simpler, but fast access to specific tiles is not enabled

Engineering Contradiction:

Improvefast access to tilesVSAvoidmedia file structure

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-calculating and storing tile position information in the media file metadata during the coding phase. The tile offset values are computed and saved in advance, so that during playback, the decoder can directly retrieve the position information without performing full picture decoding. This resolves the contradiction by enabling fast access to tiles (improving ease of operation) while using a relatively simple metadata structure (managing device complexity).

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent adds a new dimension to the media file structure by introducing metadata that stores tile offset information. This metadata layer provides fast access to tile positions without fundamentally changing the core picture coding structure. By adding information in a separate metadata dimension rather than complicating the picture data structure itself, the system achieves fast tile access while keeping the overall structure relatively simple.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Data Source

PatentUS10448031B2Method of generating media file and storage medium storing media file generation program

Publication Date: 2019.10.15 CANON KK

AI summary

A method of generating a media file using a media file format in which a set of pictures including one or more pictures is coded and stored such that each picture is divided, in coding order, into two or more slices, and coded data of each slice is stored as NAL unit data, the method comprising: dividing each slice into two or more rectangular-shaped tiles and coding the two or more rectangular-shaped tiles; and providing a slice index box in the media file format such that a value indicating an ordinal position of each slice to which each tile belongs in each picture is described in the slice index box.