# US 12,603,983 B2 · Apparatus and method for generating object-based stereoscopic images

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

- Number: US 12,603,983 B2
- Document ID: US12603983B2
- Status: Granted
- Kind: B2
- Application number: 19/010,361
- Filed: 2025-01-06
- Published: 2025-07-31
- Granted: 2026-04-14
- Earliest claimed: 2024-01-29
- Assignees: Electronics and Telecommunications Research Institute
- Applicants: Electronics and Telecommunications Research Institute
- Inventors: Seongjin Park; Do Hyung Kim
- Web page: https://patentgrep.com/patent/US12603983B2

## CPC

- H04N 13/282: for generating image signals corresponding to three or more geometrical viewpoints, e.g. multi-view systems
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

According to an embodiment of the present disclosure, a method of generating a stereoscopic video includes: obtaining video data; extracting at least one multiple viewpoint video data from the video data; identifying an object associated with the video data; generating tracking information for the object based on the at least one multiple viewpoint video data; and performing rendering on the object based on a depth value corresponding to the tracking information.

## Claims

### Claim 1

A method of generating a stereoscopic video of a stereoscopic video generating device, the method comprising:
obtaining video data;
extracting at least one multiple viewpoint video data from the obtained video data;
identifying an object associated with the obtained video data;
generating tracking information for the object based on the at least one multiple viewpoint video data;
rendering the object based on a depth value corresponding to the tracking information,
generating a local stack corresponding to each frame of the video data by implementing a plenoptic refocus technology,
extracting at least one multiple viewpoint video data based on the generated local stack, and
applying a predetermined tracking algorithm to the at least one multiple viewpoint video data to identify multiple viewpoint videos in which the object is clearly displayed among the at least one multiple viewpoint video data.

### Claim 2 (depends on 1)

The method of claim 1, wherein the video data includes plenoptic video data including color information for a plurality of views.

### Claim 3 (depends on 1)

The method of claim 1, wherein the at least one multiple viewpoint video data corresponds to at least one local video included in a local stack extracted from a plenoptic video.

### Claim 4 (depends on 1)

The method of claim 1, further comprising:
obtaining designated object information according to a user's selection.

### Claim 5 (depends on 1)

The method of claim 1, comprising:
identifying the type of display from which the object is output; and
selecting a rendering method for the object based on the type.

### Claim 6 (depends on 5)

The method of claim 5, comprising:
when the type of display is a personalized display, rendering a left binary viewpoint video and a right binary viewpoint video, and
when the type of display is a general-purpose display, rendering a multiple viewpoint video.

### Claim 7

A stereoscopic video generating apparatus, comprising:
a camera configured to receive video data;
a processor configured to generate output videos using the video data; and
a display configured to display the output videos,
wherein the processor is further configured to:
receive the video data,
extract at least one multiple viewpoint video data from the received video data,
identify the object associated with the received video data,
generate tracking information for the object based on the at least one multiple viewpoint video data, and
generate the output video by rendering the object based on a depth value corresponding to the tracking information,
wherein the processor is further configured to:
generate a local stack corresponding to each frame of the video data by implementing a plenoptic refocus technology,
extract at least one multiple viewpoint video data based on the generated local stack, and
apply a predetermined tracking algorithm to the at least one multiple viewpoint video data to identify multiple viewpoint videos in which the object is clearly displayed among the at least one multiple viewpoint video data.

### Claim 8 (depends on 7)

The apparatus of claim 7, wherein the video data includes plenoptic video data including color information for multiple viewpoints.

### Claim 9 (depends on 7)

The apparatus of claim 7, wherein the at least one multiple viewpoint video data corresponds to at least one local video included in a local stack extracted from a plenoptic video.

### Claim 10 (depends on 7)

The apparatus of claim 7, wherein the processor is further configured to receive designated object information according to a user's selection.

### Claim 11 (depends on 7)

The apparatus of claim 7, wherein the processor identifies the type of display from which the object is output; and selects a rendering method for the object based on the type.

### Claim 12 (depends on 11)

The apparatus of claim 11, wherein
when the type of display is a personalized display, the processor is further configured to render a left binary viewpoint video and a right binary viewpoint video, and
when the type of display is a general-purpose display, the processor is further configured to render a multiple viewpoint video.