Invention Grant
- Patent Title: High resolution 3D point clouds generation based on CNN and CRF models
-
Application No.: US15641113Application Date: 2017-07-03
-
Publication No.: US10671082B2Publication Date: 2020-06-02
- Inventor: Yu Huang , Hsien-Ting Cheng , Jun Zhu , Weide Zhang
- Applicant: Baidu USA LLC
- Applicant Address: US CA Sunnyvale
- Assignee: BAIDU USA LLC
- Current Assignee: BAIDU USA LLC
- Current Assignee Address: US CA Sunnyvale
- Agency: Womble Bond Dickinson (US) LLP
- Main IPC: G05D1/02
- IPC: G05D1/02 ; H04N7/18 ; H04N5/232 ; G06T3/40 ; G06T7/55 ; G06T7/521 ; G06T11/60 ; G05D1/00 ; G01S17/89 ; G06N3/04 ; G01S7/48 ; H04N13/128 ; H04N13/271 ; H04N13/254 ; H04N13/00 ; G06T5/50 ; G06T5/00 ; H04N5/225 ; G06N7/00 ; G01S17/86 ; G01S17/931 ; G06N3/08

Abstract:
In one embodiment, a method or system generates a high resolution 3-D point cloud to operate an autonomous driving vehicle (ADV) from a low resolution 3-D point cloud and camera-captured image(s). The system receives a first image captured by a camera for a driving environment. The system receives a second image representing a first depth map of a first point cloud corresponding to the driving environment. The system determines a second depth map by applying a convolutional neural network model to the first image. The system generates a third depth map by applying a conditional random fields model to the first image, the second image and the second depth map, the third depth map having a higher resolution than the first depth map such that the third depth map represents a second point cloud perceiving the driving environment surrounding the ADV.
Public/Granted literature
- US20190004535A1 HIGH RESOLUTION 3D POINT CLOUDS GENERATION BASED ON CNN AND CRF MODELS Public/Granted day:2019-01-03
Information query