# US 8,787,444 B2 · Differential coding of intra directions (DCIC)

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 8,787,444 B2
- Document ID: US8787444B2
- Status: Granted
- Kind: B2
- Application number: 13/176,132
- Filed: 2011-07-05
- Published: 2012-01-19
- Granted: 2014-07-22
- Earliest claimed: 2010-07-16
- Assignees: Sony Corporation
- Applicants: Ehsan Maani; Wei Liu; Ali Tabatabai
- Inventors: Ehsan Maani; Wei Liu; Ali Tabatabai
- Web page: https://patentgrep.com/patent/US8787444B2

## CPC

- H04N 19/463: by compressing encoding parameters before transmission
- H04N 19/593: involving spatial prediction techniques
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

Spatial domain directional intra prediction has been shown to be very effective to remove the correlation between the pixels in the current block and reconstructed neighbors. In AVC, 8 directional prediction modes (plus the DC prediction mode) are defined. The prediction mode number is signaled to the decoder using a simple predictive coding method. The previous intra prediction methods have two major disadvantages: lack of precision for arbitrary directional patterns and accuracy lack of accuracy to exploit geometric dependency between blocks. To address these issues, a new method accurately predicts the intra directions from reconstructed neighboring pixels and differentially encode the intra directions. This allows a more precise directional prediction without the significant increase in the cost for transmitting the side information.

## Claims

### Claim 1

A method of encoding an image programmed in a controller in a device comprising:
a. applying operators to neighboring pixels of a block to obtain an image gradient vector for each of the pixels;
b. determining an edge direction for each of the pixels by taking a perpendicular direction to each of the gradient vectors; and
c. predicting a best direction using the edge direction for each of the pixels; wherein the best direction is adjusted by a direction adjustment, wherein the direction adjustment is implemented using discrete values with a quantization step size, wherein the quantization step size implements a confidence measure, wherein the higher the confidence measure, the smaller the quantization step size.

### Claim 2 (depends on 1)

The method of claim 1 wherein the best direction is used in intra coding in encoding the image.

### Claim 3 (depends on 1)

The method of claim 1 wherein predicting further comprises deriving the best direction using a summation of the square of projections of direction vectors onto a unit vector.

### Claim 4 (depends on 1)

The method of claim 1 wherein predicting further comprises deriving the best direction by calculating an eigenvector using a matrix and then using a lookup table to convert the eigenvector to the best direction.

### Claim 5 (depends on 1)

The method of claim 1 wherein the confidence measure is compared to a threshold for choosing between two pre-defined step size values for the quantization step size.

### Claim 6 (depends on 1)

The method of claim 1 wherein the device is selected from the group consisting of a personal computer, a laptop computer, a computer workstation, a server, a mainframe computer, a handheld computer, a personal digital assistant, a cellular/mobile telephone, a smart appliance, a gaming console, a digital camera, a digital camcorder, a camera phone, an iPod®/iPhone, a video player, a DVD writer/player, a Blu-ray® writer/player, a television and a home entertainment system.

### Claim 7

An apparatus for encoding an image programmed in a controller in a device comprising:
a. an image gradient vector module for applying operators to neighboring pixels for obtaining an image gradient vector for each of the pixels;
b. an edge detection module for determining an edge direction for each of the pixels by taking a perpendicular direction to each of the gradient vectors; and
c. a best direction module for predicting a best direction using the edge direction of each of the pixels, wherein the best direction is adjusted by a direction adjustment, wherein the direction adjustment is implemented using discrete values with a quantization step size, wherein the quantization step size implements a confidence measure, wherein the higher the confidence measure, the smaller the quantization step size.

### Claim 8 (depends on 7)

The apparatus of claim 7 wherein the best direction is used in intra coding in encoding the image.

### Claim 9 (depends on 7)

The apparatus of claim 7 wherein the best direction module is further for deriving the best direction using a summation of the square of projections of direction vectors onto a unit vector.

### Claim 10 (depends on 7)

The apparatus of claim 7 wherein predicting further comprises deriving the best direction by calculating an eigenvector using a matrix and then using a lookup table to convert the eigenvector to the best direction.

### Claim 11 (depends on 7)

The apparatus of claim 7 wherein the confidence measure is compared to a threshold for choosing between two pre-defined step size values for the quantization step size.

### Claim 12 (depends on 7)

The apparatus of claim 7 wherein the device is selected from the group consisting of a personal computer, a laptop computer, a computer workstation, a server, a mainframe computer, a handheld computer, a personal digital assistant, a cellular/mobile telephone, a smart appliance, a gaming console, a digital camera, a digital camcorder, a camera phone, an iPod®/iPhone, a video player, a DVD writer/player, a Blu-ray® writer/player, a television and a home entertainment system.

### Claim 13

A device comprising:
a. a memory for storing an application, the application for:
i. applying operators to neighboring pixels of a block to obtain an image gradient vector for each of the pixels;
ii. determining an edge direction for each of the pixels by taking a perpendicular direction to each of the gradient vectors; and
iii. predicting a best direction using the edge direction for each of the pixels, wherein the best direction is adjusted by a direction adjustment, wherein the direction adjustment is implemented using discrete values with a quantization step size, wherein the quantization step size implements a confidence measure, wherein the higher the confidence measure, the smaller the quantization step size; and
b. a processing component coupled to the memory, the processing component configured for processing the application.

### Claim 14 (depends on 13)

The device of claim 13 wherein the best direction is used in intra coding in encoding the image.

### Claim 15 (depends on 13)

The device of claim 13 wherein predicting further comprises deriving the best direction using a summation of the square of projections of direction vectors onto a unit vector.

### Claim 16 (depends on 13)

The device of claim 13 wherein predicting further comprises deriving the best direction by calculating an eigenvector using a matrix and then using a lookup table to convert the eigenvector to the best direction.

### Claim 17 (depends on 13)

The device of claim 13 wherein the confidence measure is compared to a threshold for choosing between two pre-defined step size values for the quantization step size.

### Claim 18

A method of encoding an image programmed in a controller in a device comprising:
a. differentially encoding is implemented for a direction if the direction is close to suggested directions of neighbors of the direction; and
b. otherwise, using a simple fixed length code to signal the direction of the direction, wherein differentially encoding further comprises:
i. applying operators to neighboring pixels of a block to obtain an image gradient vector for each of the pixels;
ii. determining an edge direction for each of the pixels by taking a perpendicular direction to each of the gradient vectors; and
iii. predicting a best direction using the edge direction for each of the pixels, wherein the best direction is adjusted by a direction adjustment, wherein the direction adjustment is implemented using discrete values with a quantization step size, wherein the quantization step size implements a confidence measure, wherein the higher the confidence measure, the smaller the quantization step size.

### Claim 19 (depends on 18)

The method of claim 18 wherein the best direction is used in intra coding in encoding the image.

### Claim 20 (depends on 18)

The method of claim 18 wherein predicting further comprises deriving the best direction using a summation of the square of projections of direction vectors onto a unit vector.

### Claim 21 (depends on 18)

The method of claim 18 wherein predicting further comprises deriving the best direction by calculating an eigenvector using a matrix and then using a lookup table to convert the eigenvector to the best direction.

### Claim 22 (depends on 18)

The method of claim 18 wherein the confidence measure is compared to a threshold for choosing between two pre-defined step size values for the quantization step size.

### Claim 23 (depends on 18)

The method of claim 18 wherein the device is selected from the group consisting of a personal computer, a laptop computer, a computer workstation, a server, a mainframe computer, a handheld computer, a personal digital assistant, a cellular/mobile telephone, a smart appliance, a gaming console, a digital camera, a digital camcorder, a camera phone, an iPod®/iPhone, a video player, a DVD writer/player, a Blu-ray® writer/player, a television and a home entertainment system.