Patents by Inventor Kwang Hoon Sohn
Kwang Hoon Sohn has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20220051433Abstract: The present disclosure provides an apparatus for cylindrical convolutional neural network operation, and an apparatus and method for object recognition and viewpoint estimation using the same. The present disclosure uses a cylindrical convolutional network to recognize objects in the image and to determine the viewpoint, which cylindrical convolutional network performs a convolution operation while sliding the input image in the direction of rotation of the cylindrical kernel and extracts a plurality of view-specific feature vectors according to angular unit intervals, to identify and recognize objects, as well as to determine the viewpoint from which the objects were photographed.Type: ApplicationFiled: September 8, 2021Publication date: February 17, 2022Inventors: Kwang Hoon SOHN, Sung Hun JOUNG
-
Patent number: 10289933Abstract: A method and device for transforming 2D images into 3D are disclosed. The disclosed device includes a dictionary storage unit configured to store a word-depth gradient dictionary; a color patch obtainer unit configured to obtain color patches from an input image; a matching word search unit configured to transform each of the color patches obtained by the color patch obtainer unit into a SIFT descriptor form and search for words closest to the SIFT descriptors of the obtained color patches from among the words of the word-depth gradient dictionary; a matching depth gradient obtainer unit configured to obtain depth gradient information of the words matching the obtained color patches from the word-depth gradient dictionary; and a depth map generation unit configured to compute a depth from the obtained matching depth gradient for each of the obtained color patches and generate a depth map.Type: GrantFiled: August 24, 2016Date of Patent: May 14, 2019Assignee: INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITYInventors: Kwang Hoon Sohn, Sun Ok Kim
-
Patent number: 9601842Abstract: The present invention discloses a matable and dematable electrical connecting structure characterized by comprising: a female coupling member having a first connecting portion; a male coupling member having a second connecting portion; and a connecting unit coupling the female coupling member and the male coupling member and electrically connecting the first and second connecting portions, wherein the connecting unit includes an inner conductive material which is electrically connected to the first connecting portion and is provided on the inner wall of an insert hole formed in the female connecting member, a column including a conductive material which is electrically connected to the second connecting portion, protruding from the male connecting member, and can be inserted in the insert hole, and one or more elastic pin including a surface of a conductive material which is extending in an outward direction from the column and elastically contacting the inner conductive material.Type: GrantFiled: August 21, 2013Date of Patent: March 21, 2017Assignees: Unid Co., Ltd., OCI CorporationInventors: Dong Su Lee, Chong Kwang Yoon, Hyeong Seok Lee, Kwang Hoon Sohn
-
Publication number: 20170061247Abstract: A method and device for transforming 2D images into 3D are disclosed. The disclosed device includes a dictionary storage unit configured to store a word-depth gradient dictionary; a color patch obtainer unit configured to obtain color patches from an input image; a matching word search unit configured to transform each of the color patches obtained by the color patch obtainer unit into a SIFT descriptor form and search for words closest to the SIFT descriptors of the obtained color patches from among the words of the word-depth gradient dictionary; a matching depth gradient obtainer unit configured to obtain depth gradient information of the words matching the obtained color patches from the word-depth gradient dictionary; and a depth map generation unit configured to compute a depth from the obtained matching depth gradient for each of the obtained color patches and generate a depth map.Type: ApplicationFiled: August 24, 2016Publication date: March 2, 2017Inventors: Kwang Hoon SOHN, Sun Ok KIM
-
Patent number: 9483699Abstract: An apparatus and a method for detecting traffic lanes in real time are disclosed. The disclosed real-time lane detection apparatus may include: a candidate area establisher unit configured to establish as a candidate area for lane detection an area having an intensity value which corresponds to the intensity of a traffic lane marking from among the intensity values in a color space of a color image; and a lane-marking determiner unit configured to determine a traffic lane marking from the established candidate area by using a line component of the candidate area. According to the present invention, traffic lanes can be detected accurately and quickly, even in environments where the lighting conditions of the road vary.Type: GrantFiled: March 7, 2014Date of Patent: November 1, 2016Assignee: Industry-Academic Cooperation Foundation, Yonsei UniversityInventors: Kwang-Hoon Sohn, Jong-In Son
-
Patent number: 9443283Abstract: A content retargeting method and apparatus are disclosed. An embodiment of the invention provides a content retargeting method that includes: dividing an original content into a grid having M rows×N columns of quads; computing degrees of importance of the divided quads; and scaling the quads based on the computed degrees of importance, the quads axis-aligned by rows or columns.Type: GrantFiled: September 17, 2014Date of Patent: September 13, 2016Assignee: INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITYInventors: Kwang-Hoon Sohn, Sung-Hwan Choi, Bum Sub Ham
-
Publication number: 20160149318Abstract: The present invention discloses a matable and dematable electrical connecting structure characterized by comprising: a female coupling member having a first connecting portion; a male coupling member having a second connecting portion; and a connecting unit coupling the female coupling member and the male coupling member and electrically connecting the first and second connecting portions, wherein the connecting unit includes an inner conductive material which is electrically connected to the first connecting portion and is provided on the inner wall of an insert hole formed in the female connecting member, a column including a conductive material which is electrically connected to the second connecting portion, protruding from the male connecting member, and can be inserted in the insert hole, and one or more elastic pin including a surface of a conductive material which is extending in an outward direction from the column and elastically contacting the inner conductive material.Type: ApplicationFiled: August 21, 2013Publication date: May 26, 2016Inventors: Dong Su LEE, Chong Kwang YOON, Hyeong Seok LEE, Kwang Hoon SOHN
-
Patent number: 9225967Abstract: Provided is an image processing apparatus, method and computer-readable medium. When encoding a multi-view color and depth video, a motion vector calculator of the image processing apparatus may calculate a block motion vector with respect to a first viewpoint color image. A motion vector borrowing unit may borrow, as a block motion vector with respect to a second viewpoint color image, a block motion vector with respect to the first viewpoint color image based on a disparity calculated from a depth video.Type: GrantFiled: January 21, 2011Date of Patent: December 29, 2015Assignee: INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITYInventors: Kwang Hoon Sohn, Seok Lee, Ho Cheon Wey, Jung Dong Seo, Dong Hyun Kim
-
Publication number: 20150348230Abstract: A content retargeting method and apparatus are disclosed. An embodiment of the invention provides a content retargeting method that includes: dividing an original content into a grid having M rows×N columns of quads; computing degrees of importance of the divided quads; and scaling the quads based on the computed degrees of importance, the quads axis-aligned by rows or columns.Type: ApplicationFiled: September 17, 2014Publication date: December 3, 2015Inventors: Kwang-Hoon SOHN, Sung-Hwan CHOI, Bum Sub HAM
-
Patent number: 9098936Abstract: An apparatus for enhancing a stereoscopic image may include: a color relationship extraction unit, which extracts color relationships between a plurality of first coordinates in a 3-dimensional color space for a first image and second coordinates in a 3-dimensional color space for a second image corresponding to the plurality of first coordinates; a color relationship correction unit, which corrects a color relationship for any one first coordinate from among the plurality of first coordinates based on a color relationship of at least one first coordinate existing within a particular distance from the any one first coordinate; and a color value transformation unit, which transforms a color value of the first image by using the corrected color relationship of the any one first coordinate. The invention provides the advantage of accurately correcting color imbalance between the left image and right image forming a stereoscopic image.Type: GrantFiled: December 20, 2012Date of Patent: August 4, 2015Assignee: Industry-Academic Cooperation Foundation, Yonsei UniversityInventors: Kwang-Hoon Sohn, Hyoung-Chul Shin
-
Publication number: 20150189255Abstract: A stereo camera is provided. A stereo camera according to an embodiment of the invention includes a first camera, which includes an image sensor of red, green, and blue color channels, and a second camera, which includes an image sensor that has at least some of the channels from at least one of red, green, and blue color channels replaced with infrared (IR) channels.Type: ApplicationFiled: December 30, 2014Publication date: July 2, 2015Inventors: Kwang-Hoon SOHN, Seung Chul RYU, Seung Ryong KIM, Jun Hyung KIM
-
Patent number: 8855210Abstract: Provided is a method and apparatus for encoding and decoding multi-view video data. The encoding method includes: decomposing each view image of the multi-view video into an overlapped region and a non-overlapped region, the overlapped region being overlapped with other view image and the non-overlapped region not being overlapped with other view image; generating a stitched image by combining the non-overlapped region of each view image and a middle view image; encoding the stitched image using a first encoding algorithm; and encoding the overlapped region of each view image, using a second encoding algorithm. Further, the decomposing step includes the steps of estimating disparity information for each view image, based on a predetermined view image; and decomposing each view image into said overlapped region and said non-overlapped region using the estimated disparity information.Type: GrantFiled: May 17, 2012Date of Patent: October 7, 2014Assignees: Electronics and Telecommunications Research Institute, Industry-Academic Cooperation Foundation, Yonsei UniversityInventors: Suk Hee Cho, Kug Jin Yun, Dae Hee Kim, Chung Hyun Ahn, Soo In Lee, Kwang Hoon Sohn, Yong Tae Kim
-
Publication number: 20140185879Abstract: An apparatus and a method for detecting traffic lanes in real time are disclosed. The disclosed real-time lane detection apparatus may include: a candidate area establisher unit configured to establish as a candidate area for lane detection an area having an intensity value which corresponds to the intensity of a traffic lane marking from among the intensity values in a color space of a color image; and a lane-marking determiner unit configured to determine a traffic lane marking from the established candidate area by using a line component of the candidate area. According to the present invention, traffic lanes can be detected accurately and quickly, even in environments where the lighting conditions of the road vary.Type: ApplicationFiled: March 7, 2014Publication date: July 3, 2014Applicant: Industry-Academic Cooperation Foundation, Yonsei UniversityInventors: Kwang-Hoon Sohn, Jong-In Son
-
Publication number: 20140078136Abstract: Disclosed are an apparatus and method for synthesizing an intermediate view image and a recording medium thereof. The disclosed apparatus for synthesizing an intermediate view image may include: a probability information generating part configured to generate information related to a matching probability between a first image and a second image; and a synthesizing part configured to synthesize an intermediate view image by applying a weight according to the matching probability to the first image and second image. Certain embodiments of the invention provide the advantage of enabling enhanced image quality when synthesizing an intermediate view image without requiring post-processing operations such as for processing occlusions and for filling holes.Type: ApplicationFiled: May 3, 2013Publication date: March 20, 2014Applicant: Industry-Academic Cooperation Foundation, Yonsei UniversityInventors: Kwang-Hoon Sohn, Bum Sub Ham
-
Publication number: 20140002599Abstract: Disclosed are a competition-based multiview video encoding/decoding device and a method thereof. The competition-based multiview video encoding/decoding device can improve encoding efficiency by determining a prediction vector with the best encoding performance through an extraction of a spatial prediction vector, a time prediction vector, and a viewpoint prediction vector corresponding to a current block.Type: ApplicationFiled: January 6, 2012Publication date: January 2, 2014Applicant: SAMSUNG ELECTRONICS CO., LTD.Inventors: Jin Young Lee, Dong Hyun Kim, Seung Chul Ryu, Jung Dong Seo, Kwang Hoon Sohn, Ho Cheon Wey
-
Publication number: 20120224623Abstract: Provided is a method and apparatus for encoding and decoding multi-view video data. The encoding method includes: decomposing each view image of the multi-view video into an overlapped region and a non-overlapped region, the overlapped region being overlapped with other view image and the non-overlapped region not being overlapped with other view image; generating a stitched image by combining the non-overlapped region of each view image and a middle view image; encoding the stitched image using a first encoding algorithm; and encoding the overlapped region of each view image, using a second encoding algorithm. Further, the decomposing step includes the steps of estimating disparity information for each view image, based on a predetermined view image; and decomposing each view image into said overlapped region and said non-overlapped region using the estimated disparity information.Type: ApplicationFiled: May 17, 2012Publication date: September 6, 2012Applicants: INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITY, ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTEInventors: Suk Hee CHO, Kug Jin YUN, Dae Hee KIM, Chung Hyun AHN, Soo In LEE, Kwang Hoon SOHN, Yong Tae KIM
-
Publication number: 20120189060Abstract: An apparatus and method for encoding and de1/5/2012coding motion information and disparity information are provided. The apparatus may extract a vector from peripheral blocks of a current block, which is a vector of the same type as a vector used for predicting and compensating the current block, and perform prediction and compensation of the current block using the extracted vector.Type: ApplicationFiled: January 18, 2012Publication date: July 26, 2012Applicants: INDUSTRY-ACADEMIC COOPERATION FOUNDATION, YONSEI UNIVERSITY, SAMSUNG ELECTRONICS CO., LTD.Inventors: Jin Young LEE, Ho Cheon WEY, Kwang Hoon SOHN, Jung Dong SEO, Seung Chul RYU, Dong Hyun KIM
-
Patent number: 8204133Abstract: Provided is a method and apparatus for encoding and decoding multi-view video data. The encoding method includes: decomposing each view image of the multi-view video into an overlapped region and a non-overlapped region, the overlapped region being overlapped with other view image and the non-overlapped region not being overlapped with other view image; generating a stitched image by combining the non-overlapped region of each view image and a middle view image; encoding the stitched image using a first encoding algorithm; and encoding the overlapped region of each view image, using a second encoding algorithm. Further, the decomposing step includes the steps of estimating disparity information for each view image, based on a predetermined view image; and decomposing each view image into said overlapped region and said non-overlapped region using the estimated disparity information.Type: GrantFiled: October 10, 2005Date of Patent: June 19, 2012Assignees: Electronics and Telecommunications Research Institute, Industry-Academic Cooperation Foundation, Yonsei UniversityInventors: Suk Hee Cho, Kug Jin Yun, Dae Hee Kim, Chung Hyun Ahn, Soo In Lee, Kwang Hoon Sohn, Yong Tae Kim
-
Publication number: 20110211638Abstract: Provided is an image processing apparatus, method and computer-readable medium. When encoding a multi-view color and depth video, a motion vector calculator of the image processing apparatus may calculate a block motion vector with respect to a first viewpoint color image. A motion vector borrowing unit may borrow, as a block motion vector with respect to a second viewpoint color image, a block motion vector with respect to the first viewpoint color image based on a disparity calculated from a depth video.Type: ApplicationFiled: January 21, 2011Publication date: September 1, 2011Applicants: Samsung Electronics Co., Ltd., Industry-Academic Cooperation Foundation, Yonsei UniversityInventors: Kwang Hoon Sohn, Seok Lee, Ho Cheon Wey, Jung Dong Seo, Dong Hyun Kim
-
Publication number: 20090103619Abstract: A method of coding/decoding a multiview sequence and display method thereof are disclosed, by which multiview sequence data can be efficiently coded and decoded. A multiview sequence coding method according to the present invention includes a step of generating a bit stream by encoding a plurality of pictures acquired from a plurality of views, wherein the bit stream includes view information for each of a plurality of the pictures and wherein the view information is information designating that the corresponding picture corresponds to which view among a plurality of the views. Accordingly, the multiview sequence is encoded to be selectively decoded for display.Type: ApplicationFiled: June 24, 2005Publication date: April 23, 2009Applicant: LG ELECTRONICS INC.Inventors: Kwang Hoon Sohn, Jeong Sun Lim