Vehicle behavior prediction method and device, equipment and medium

A prediction method and vehicle technology, applied in the direction of prediction, image data processing, 3D image processing, etc., can solve the problems that sub-module uncertainty cannot be effectively compensated, affect the performance of the scheme, and waste computing resources, so as to reduce computing power. The effect of resource consumption, improving calculation speed, and ensuring prediction accuracy

Active Publication Date: 2021-10-22
CHINA FIRST AUTOMOBILE
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, each module in this type of solution operates independently, so the uncertainty of its sub-modules cannot be effectively compensated during the propaga

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
  • Vehicle behavior prediction method and device, equipment and medium
  • Vehicle behavior prediction method and device, equipment and medium
  • Vehicle behavior prediction method and device, equipment and medium

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0036] Example one

[0037] figure 1 A flowchart of a method for forecasting the behavior of a vehicle according to a first embodiment of the invention, applicable to embodiments of the unmanned vehicle behavior prediction technology of the present embodiment. The method may vehicle behavior provided by the embodiment of the present invention performs prediction means, which means may be software and / or hardware implemented in the manner employed, and can be integrated on the electronic device.

[0038] Specifically, such as figure 1 , The vehicle behavior prediction method according to an embodiment of the present invention, may comprise the steps of:

[0039] S110, obtaining the three-dimensional point cloud data acquired in-vehicle sensors.

[0040] Wherein the three-dimensional point cloud data acquired by the laser sensors radar-based vehicle. Three-dimensional point cloud data is not RGB image data, comprising in addition to the collection position, but also contains a we...

Example Embodiment

[0058] Example 2

[0059] figure 2 A flowchart of a vehicle behavior prediction method according to a second embodiment of the present invention, the method further optimized based on the above-described embodiments, specific circumstances of how to extract the spatial dimension feature presentation.

[0060] Specifically, such as figure 2 As shown, the method includes:

[0061] S210, obtaining the three-dimensional point cloud data acquired in-vehicle sensors.

[0062] Wherein the three-dimensional point cloud data acquired by the laser sensors radar-based vehicle. Three-dimensional point cloud data is not RGB image data, comprising in addition to the collection position, but also contains a wealth of information, such as color information and intensity information.

[0063] In order to speed up calculations can be simplified processing of input data. After acquiring the three-dimensional point cloud data acquired in-vehicle sensor, based on three-dimensional voxel mesh, three-d...

Example Embodiment

[0076] Example three

[0077] image 3 It is a network configuration diagram of the vehicle behavior model according to a third embodiment of the present application forecast. Embodiments of the present application on the basis of the above technical solutions of the embodiments, there is provided a preferred embodiment.

[0078] Implemented on the basis of the present embodiment, the laser radar sensor on the main vehicle, and thus the conventional target detection algorithms, the input data is 3D point cloud data is not RGB image data. In the present embodiment, the input data is a simplified process, specific process can be divided into two parts: 1) a three-dimensional voxel grid based on the three-dimensional point cloud data quantizing said quantized to obtain point cloud data; 2) by two coding the quantized data values ​​of point cloud data, i.e. if the grid in the original input data memory, it is set to 1, and 0 otherwise. Since the input data is a simplified process to o...

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 discloses a vehicle behavior prediction method, apparatus and device, and a medium. The method comprises the steps of obtaining three-dimensional point cloud data collected by a vehicle-mounted sensor; based on the vehicle behavior prediction model, extracting spatial dimension features in spatial dimensions; based on a vehicle behavior prediction model, determining a compressed four-dimensional tensor according to the three-dimensional point cloud data; performing group convolution on the compressed four-dimensional tensor, and performing feature extraction on a time dimension to obtain a simplified feature map; according to the simplified feature map, performing fusion on group convolution layer results, and determining time dimension features; fusing the spatial dimension feature and the time dimension feature to obtain a fused feature; and predicting the motion state of the vehicle in the three-dimensional point cloud data according to the fusion features. According to the technical scheme, detection, tracking and behavior prediction are processed in parallel, consumption of computing resources is reduced while the prediction precision is guaranteed, the input data are compressed, and the computing speed is increased.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of unmanned driving, and in particular to a vehicle behavior prediction method, device, equipment and medium. Background technique [0002] Vehicle driving behavior prediction mainly serves applications related to the safety of the Internet of Vehicles, such as vehicle anti-collision monitoring at intersections. [0003] In the research field of unmanned driving technology at this stage, most methods divide the problem into four sub-modules, namely target detection, trajectory tracking, target behavior prediction and planning decision-making. The relationship between each module is: the output of the detection module is as The input of the tracking module, and then input the target motion curve obtained by the tracking module to the behavior prediction module, and finally output the decision signal. [0004] However, each module in this type of solution operates independently, so 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
IPC IPC(8): G06Q10/04G06T15/00G06K9/62
CPCG06Q10/04G06T15/00G06F18/253
Inventor 王祎男曹容川张天奇关瀛洲
Owner CHINA FIRST AUTOMOBILE
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