Webserver-based video content production system and method for printing holographic stereogram photo

Patent metadata

Field	Value
Publication number	US-12405573-B2
Application number	US-202217660103-A
Country	US
Kind code	B2
Filing date	Apr 21, 2022
Priority date	Dec 27, 2021
Publication date	Sep 2, 2025
Grant date	Sep 2, 2025

Abstract

Provided are a webserver-based video content production system and method for printing a holographic stereogram photo. According to the webserver-based video content production system and method, an encoding process is performed on image data under a web environment to prevent data loss in a network communication process, and then the encoded image data is transmitted to a webserver. The webserver decodes the encoded image data and performs a second-half task (postprocessing) on the decoded image data using deep learning such that a high-quality hologram photo can be printed within a relatively short time without any unnecessary process.

First claim

What is claimed is:

1. A webserver-based video content production method for printing a holographic stereogram photo, the webserver-based video content production method comprising:
providing image data acquired by imaging a subject through a user terminal to a web;
providing the received image data to a webserver;
performing a second-half task on the received image data using deep learning to produce video content; and
providing the produced video content to the user terminal through the web; and
wherein the production of the video content comprises:
recognizing a face in all frames of the image data received from the web;
deleting one or more frames with closed eyes among frames of the image data in which the face is recognized;
only extracting frames having a certain variation less than a threshold from the frames of the image data, in which the one or more frames with closed eyes has been deleted, to stabilize images and equalize the frames of the image data;
separating the subject and a background in the extracted frames having the certain variation less than the threshold; and
performing postprocessing on the images in which the subject and the background have been separated to produce video content; and
wherein the deleting of the one or more frames with closed eyes comprises:
connecting landmarks of a region, which corresponds to an eye among facial landmarks representing major features of the face, with lines and then generating the region corresponding to the eye;
calculating an area of the generated region corresponding to the eye; and
when the calculated area is a set threshold or less, recognizing one or more corresponding frames as frames with closed eyes and deleting the one or more corresponding frames.

2. The webserver-based video content production method of claim 1 , wherein the image data is any one of image or video data which is acquired by the user terminal imaging the subject through a camera or webcam provided in the user terminal after accessing the web and image or video data which is captured in advance through the camera or webcam provided in the user terminal and stored in a memory.

3. The webserver-based video content production method of claim 1 , wherein, when the user terminal accesses a webpage, the web provides description for hologram photography to the user terminal and supports an imaging assistance tool.

4. The webserver-based video content production method of claim 1 , wherein the image data is encoded in Base64 format and then transmitted to the webserver.

5. The webserver-based video content production method of claim 1 , wherein the recognizing of the face comprises recognizing the face in the frames of the image data using a Haar cascade classifier to select frames with a front of the face.

6. The webserver-based video content production method of claim 1 , wherein the stabilization of the images and the equalization of the frames comprise:
calculating a variation between previous and subsequent frames of an n th frame among the frames of the image data;
when the calculated frame variation is the set threshold or more, removing a corresponding frame; and
only extracting frames having the certain variation less than the threshold.

7. The webserver-based video content production method of claim 1 , wherein the separating of the subject and the background comprises separating a foreground and the background using saliency detection based on deep learning which is a protruding object detection technique for separating an object that is most salient when an image is viewed from a background.

8. The webserver-based video content production method of claim 1 , wherein the production of the video content further comprises:
decoding the encoded image data received from the web; and
encoding the produced video content.

9. The webserver-based video content production method of claim 1 , wherein the providing of the produced video content to the user terminal further comprises providing, by the web, the video content received from the webserver to a user through the user terminal such that the user views the video content and makes a selection.

10. The webserver-based video content production method of claim 9 , wherein the providing of the produced video content to the user terminal further comprises, when any one piece of the video content is selected by the user, printing, by the web, the selected piece of video content through a holographic stereogram printer.

11. A webserver-based video content production system for printing a holographic stereogram photo, the webserver-based video content production system comprising:
a user terminal configured to image a subject and acquire image data;
a web configured to receive the image data from the user terminal; and
a webserver configured to perform a second-half task on the image data received from the web using deep learning in response to a request from the web to produce video content and then provide the video content to the web; and
wherein the webserver produces the video content through a process comprising:
recognizing a face in all frames of the image data received from the web;
deleting one or more frames with closed eyes among frames of the image data in which the face is recognized;
only extracting frames having a certain variation less than a threshold from the frames of the image data, in which the one or more frames with closed eyes has been deleted, to stabilize images and equalize the frames of the image data;
separating the subject and a background in the extracted frames having the certain variation less than the threshold; and
performing postprocessing on the images in which the subject and the background have been separated; and
wherein the deleting of the one or more frames with closed eyes comprises:
connecting landmarks of a region, which corresponds to an eye among facial landmarks representing major features of the face, with lines and then generating the region corresponding to the eye;
calculating an area of the generated region corresponding to the eye; and
when the calculated area is a set threshold or less, recognizing one or more corresponding frames as frames with closed eyes and deleting the one or more corresponding frames.

12. The webserver-based video content production system of claim 11 , wherein the image data is any one of image or video data which is acquired by the user terminal imaging the subject through a camera or webcam provided in the user terminal after accessing the web and image or video data which is captured in advance through the camera or webcam provided in the user terminal and stored in a memory.

13. The webserver-based video content production system of claim 11 , wherein, when the user terminal accesses a webpage, the web provides description for hologram photography to the user terminal and supports an imaging assistance tool.

14. The webserver-based video content production system of claim 11 , wherein the web encodes the image data in Base64 format and then transmits the encoded image data to the webserver.

15. The webserver-based video content production system of claim 11 , wherein the recognizing of the face comprises recognizing the face in the frames of the image data using a Haar cascade classifier to select frames with a front of the face.

16. The webserver-based video content production system of claim 11 , wherein the stabilization of the images and the equalization of the frames comprise:
calculating a variation between previous and subsequent frames of an n th frame among the frames of the image data;

Assignees

• Kwangwoon Univ Industry—Academic Collaboration Foundation
• Univ Kwangwoon Ind Acad Collab

Inventors

• Lee Seung Hyun
• Hwang Lee Hwan
• Kwon Soon Chul
• Park Chan Soo
• Shin Seok Yong
• Gentet Philippe
• Kim Jung Ho
• Kim Yong Jung

Classifications

• G06V10/95 Primary

  structured as a network, e.g. client-server architectures · CPC title

• G06V10/82

  using neural networks · CPC title

• G03H2226/04

  Transmission or communication means, e.g. internet protocol · CPC title

• G06V20/46

  Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames · CPC title

• G03H2001/269

  Two and more than two steps recording process · CPC title