公开(公告)号：US20190080604A1

公开(公告)日：2019-03-14

申请号：US16113043

申请日：2018-08-27

Applicant: Connaught Electronics Ltd.

Inventor： Senthil Kumar Yogamani ,  Sunil Chandra

IPC: G08G1/14 ,  G06N3/08 ,  G08G1/04 ,  G06K9/00 ,  G06K9/66

Abstract: The invention relates to a method for operating a driver assistance system (2) of a motor vehicle (1), including a) Capturing an environment (4) of the motor vehicle (1) by a capturing device (3) of the driver assistance system (2); b) Detecting an accessible freespace (6) in the captured environment (4) by a computing device (5) of the driver assistance system (2); c) Detecting and Classifying at least one object (7a-7e) in the captured environment (4) that is located at a border (8) of the freespace (6) by a neural network (9) of the driver assistance system (2); d) Assigning a part (10a-10e) of the border (8) of the freespace (6) to the detected and classified object (7a-7e); and e) Categorizing a part (11a-11e) of the freespace (6) adjacent to the part (10a-10e) of the border (8) that is assigned to the detected and classified object (7a-7e) in dependence upon the class of that classified object (7a-7e), so as to enable improved safety in driving.

公开(公告)号：US12217444B2

公开(公告)日：2025-02-04

申请号：US17784896

申请日：2020-12-08

Applicant: Connaught Electronics Ltd.

Inventor： Ciaran Hughes ,  Pantelis Ermilios ,  Matthias Lenk ,  Sunil Chandra

IPC: G06T7/579 ,  G06T7/593 ,  H04N13/221 ,  H04N13/243

Abstract: A method for forming a point cloud corresponding to the topography of an imaged environment involves acquiring an image of the environment with a camera having a WFOV lens mounted on a vehicle; changing the camera pose by an adjustment greater than a threshold; and acquiring another image of the environment with the camera at the changed pose. The images are mapped onto respective surfaces to form respective mapped images, defined by the same nonplanar geometry. One of the mapped images is divided into blocks of pixels. For each block, a depth map is formed by performing a search through the other mapped image to evaluate a disparity in the position of the location of the block of pixels in each of the mapped images. The depth map is converted into a partial point cloud corresponding to the local topography of the environment surrounding the vehicle as it moves through the environment.

公开(公告)号：US20230351625A1

公开(公告)日：2023-11-02

申请号：US17784896

申请日：2020-12-08

Applicant: Connaught Electronics Ltd.

Inventor： Ciaran Hughes ,  Pantelis Ermilios ,  Matthias Lenk ,  Sunil Chandra

IPC: G06T7/579 ,  G06T7/593 ,  H04N13/243 ,  H04N13/221

CPC classification number: G06T7/579 ,  G06T7/593 ,  H04N13/243 ,  H04N13/221 ,  G06T2207/30252 ,  G06T2207/20021 ,  G06T2207/10028

Abstract: A method for forming a point cloud corresponding to the topography of an imaged environment involves acquiring an image of the environment with a camera having a WFOV lens mounted on a vehicle; changing the camera pose by an adjustment greater than a threshold; and acquiring another image of the environment with the camera at the changed pose. The images are mapped onto respective surfaces to form respective mapped images, defined by the same nonplanar geometry. One of the mapped images is divided into blocks of pixels. For each block, a depth map is formed by performing a search through the other mapped image to evaluate a disparity in the position of the location of the block of pixels in each of the mapped images. The depth map is converted into a partial point cloud corresponding to the local topography of the environment surrounding the vehicle as it moves through the environment.

公开(公告)号：US10950127B2

公开(公告)日：2021-03-16

申请号：US16113043

申请日：2018-08-27

Applicant: Connaught Electronics Ltd.

Inventor： Senthil Kumar Yogamani ,  Sunil Chandra

IPC: G08G1/14 ,  G06K9/00 ,  G06K9/62 ,  G06K9/66 ,  G06N3/08 ,  G08G1/04

Abstract: The invention relates to a method for operating a driver assistance system (2) of a motor vehicle (1), including a) Capturing an environment (4) of the motor vehicle (1) by a capturing device (3) of the driver assistance system (2); b) Detecting an accessible freespace (6) in the captured environment (4) by a computing device (5) of the driver assistance system (2); c) Detecting and Classifying at least one object (7a-7e) in the captured environment (4) that is located at a border (8) of the freespace (6) by a neural network (9) of the driver assistance system (2); d) Assigning a part (10a-10e) of the border (8) of the freespace (6) to the detected and classified object (7a-7e); and e) Categorizing a part (11a-11e) of the freespace (6) adjacent to the part (10a-10e) of the border (8) that is assigned to the detected and classified object (7a-7e) in dependence upon the class of that classified object (7a-7e), so as to enable improved safety in driving.