Obstacle detection method based on combination of point cloud information and deep learning

A technology of obstacle detection and point cloud information, which is applied in the field of obstacle detection, can solve problems such as the single obstacle detection method of robots and the inability to fully perceive obstacles, so as to improve intelligence, reduce data processing scale, and reduce data processing scale Effect

Pending Publication Date: 2022-04-01
STATE GRID INTELLIGENCE TECH CO LTD
View PDF2 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In order to solve the above problems, this invention proposes an obstacle detection method combining point cloud information and deep learning. Inspection of various obstacles encountered by the robot to ensure the safe operation of the equipment

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Obstacle detection method based on combination of point cloud information and deep learning
  • Obstacle detection method based on combination of point cloud information and deep learning
  • Obstacle detection method based on combination of point cloud information and deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0043] A obstacle detection method combined with point cloud information and depth learning, such as figure 1 As shown, including the following steps:

[0044] The disorder detection is obtained by using binocular vision.

[0045] In the present embodiment, the specific selection of the double-eyed visual camera is an Intel's RealSense D435 depth camera, and the front side of the camera is from left to the right is the left infrared camera, an infrared enhancement transmitter, a right infrared camera, and a visible light camera.

[0046] The camera is powered and data transmission in USB3.0, and the system is accessible. And with the REALSENSE camera and depth learning technology, the Realsense camera is widely used in 3D target detection, limb behavior analysis, robot visual navigation and avoidance, virtual reality, etc. While detecting an obstacle, classify the category of the obstacle, provides a basis for subsequent waiver.

[0047] The Realsense ranging ranges from 0.2 to 10...

Embodiment 2

[0068] A obstacle detection system combined with point cloud information and depth learning, including:

[0069] The image acquisition module is configured to obtain a point cloud image containing depth information;

[0070] The pretreatment module is configured to filter the image, remove the processing of the extension point of the distribution edge, resulting in the pretreatment point cloud information;

[0071] The ground information extraction module is configured to extract the ground information in the point cloud information, and filter out;

[0072] The initial detection module is configured to perform preliminary obstacle detection based on the remaining point cloud information, and determine the grill of the obstacle;

[0073] The determination classification module is configured to determine the type of obstacle using the preliminary obstacle using the pre-training depth learning model to determine the type of obstacle.

Embodiment 3

[0075] A computer readable storage medium in which a plurality of instructions are stored, and the instructions are adapted to load and perform the steps provided by the terminal device.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a point cloud information and deep learning combined obstacle detection method. The method comprises the following steps: acquiring a point cloud image containing depth information; filtering the image, and removing outliers of a distribution edge to obtain preprocessed point cloud information; extracting and filtering ground information in the point cloud information; based on the residual point cloud information, carrying out preliminary obstacle detection, and determining a grid where an obstacle is located; judging the preliminarily detected obstacle by using a pre-trained deep learning model, and determining the type of the obstacle; the defects that an existing robot is single in obstacle detection mode and cannot comprehensively sense obstacles can be overcome, various obstacles encountered by the indoor inspection robot can be effectively detected in real time, and operation safety of equipment is guaranteed.

Description

Technical field [0001] The present invention belongs to the field of obstacle detection, and more particularly to a disorder detection method in which point cloud information is combined with depth learning. Background technique [0002] The statement of this section is merely the background technology information associated with the present invention, which is not necessarily constituted in prior art. [0003] Indoor inspection robots are dedicated robots in the secondary equipment protection room of the substation of the substation or in communication machine room, obstacle detection technology is key technologies to ensure the daily operation of robots. Current obstacles detect common sensors include laser sensors, ultrasonic sensors, and visual sensors. The laser sensor is high, but only the location information and contour information of the obstacle can be obtained, and the robot cannot help the robot fully cope with the type of obstacle and the hardware cost is high. The u...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06V20/10G06T5/00G06N3/08G06K9/62G06V10/774G06V10/82G06V10/30G06T7/55G06V10/764
Inventor 杨杰邵光亭巩方彬刘加科王贤华蔺茹徐云龙孙大庆黄倩菁朱琳亓曙光吕庆涛
Owner STATE GRID INTELLIGENCE TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap