Road barrier identification method and device

A technology for obstacle recognition and obstacles, applied in image recognition and computer fields, can solve the problems that the recognition results cannot be corrected in time, reduce the recognition accuracy of machine learning models, etc., and achieve the effect of improving the recognition accuracy

CN105957145AInactive Publication Date: 2016-09-21BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
6 Cites 53 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2016-09-21
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a road barrier identification method and device. One specific embodiment of the invention comprises the steps of: obtaining laser point cloud of a barrier object on a road; based on the laser point cloud, obtaining mark information of the barrier object; and based on the mark information, generating training samples of a machine learning model. According to the invention, different type of mark information of the barrier objects are obtained based on the laser point cloud of the barrier objects, so that the sizes, positions and angles of the different kinds of barrier objects are accurately determined; and furthermore, based on the mark information of the barrier object, the training samples of the machine learning model for identifying the sizes, positions and angles of the different kinds of barriers using the laser point cloud of the barriers as input are generated, and the identification accuracy of the machine learning model is continuously improved.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present application relates to the field of computers, in particular to the field of image recognition, especially to a method and device for road obstacle recognition. Background technique

[0002] The extensive application of computer technology in automobiles makes automobiles more and more intelligent. During the driving process of the car, the laser point cloud of obstacles on the road can be obtained by using the on-board laser scanning equipment. It is necessary to use the machine learning model to use the laser point cloud as input to identify the size, position, and angle of the obstacle.

[0003] However, since the obtained laser point clouds of all obstacle objects are mixed together, there will be situations where the machine learning model cannot correctly distinguish the overlapping obstacle objects, which will lead to wrong recognition results that cannot be corrected in time, reducing the The recognition accuracy of the machine lea...

Examples

Embodiment Construction

[0014] The application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain related inventions, rather than to limit the invention. It should also be noted that, for the convenience of description, only the parts related to the related invention are shown in the drawings.

[0015] It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and embodiments.

[0016] figure 1 An exemplary system architecture 100 to which embodiments of the method or device for identifying road obstacles of the present application can be applied is shown.

[0017] Such as figure 1 As shown, the system architecture 100 may include terminal devices...