Lane Line Data Processing Using Deep Neural Networks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for detecting and generating vehicular lane lines in high-precision maps are inefficient and labor-intensive, with manual approaches being costly and prone to errors, while automatic methods lack the precision required for mass production of high-precision maps.

Innovation Solution

A vehicular lane line data processing method utilizing a deep neural network to analyze images from industrial cameras, calculating pixel confidence, filtering candidate lane lines based on geometric characteristics, and projecting two-dimensional data into three-dimensional maps, employing technologies like deep learning and machine vision for accurate lane line recognition and reconstruction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual approach is used to draw lane lines on point cloud, then labor cost is reduced, but detection precision and efficiency deteriorate

Engineering Contradiction:
Improvelane line detection precisionVSAvoidlane line detection efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual mechanical drawing operations with an automated computer-based system that processes point cloud data and road images through algorithms to automatically generate lane line annotations, eliminating the need for manual interaction while maintaining high precision

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables the point cloud and road image data to self-generate lane line annotations through automated processing, where the computer independently performs detection, filtering, and drawing operations without human intervention, achieving both high precision and efficiency

Inventive Principle:
Principle #25Self-service

2Productivity

If automatic recognition method is used to detect lane lines, then productivity is improved, but measurement precision deteriorates

Engineering Contradiction:
Improvelane line detection efficiencyVSAvoidlane line detection precision
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent merges point cloud data processing with road image processing in a unified automated system, combining the spatial information from point clouds with the visual information from images to achieve both high precision detection and efficient automated processing

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The system uses road images as an intermediary to guide the detection process, where image-based lane line detection results serve as references to filter and validate point cloud data, improving precision while maintaining automated efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If point cloud with low resolution is used for manual drawing, then device complexity is reduced, but measurement precision deteriorates

Engineering Contradiction:
Improvelane line detection precisionVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent combines point cloud data with road image data to create a fused representation that compensates for the low resolution of point clouds, using the complementary information from images to achieve high precision lane line detection without requiring complex high-resolution point cloud acquisition systems

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3171292B1Driving lane data processing method, device, storage medium and apparatus
Publication Date: 2023.11.08 BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
  • EP3171292B1 patent drawingFigure 1
  • EP3171292B1 patent drawingFigure 2
  • EP3171292B1 patent drawingFigure 3

AI summary

The embodiments of the present disclosure disclose a vehicular lane line data processing method, apparatus, storage medium, and device. The method includes: acquiring at least two consecutive original images of a vehicular lane line and positioning data of the original images; calculating, using a deep neural network model, a pixel confidence for a conformity between a pixel characteristic in the original images and a vehicular lane line characteristic; determining an outline of the vehicular lane line from the original images and using the outline of the vehicular lane line as a candidate vehicular lane line; calculating a vehicular lane line confidence of the candidate vehicular lane line based on the pixel confidences of pixels in the candidate vehicular lane line; filtering the candidate vehicular lane line based on the vehicular lane line confidence of the candidate vehicular lane line; recognizing, for the filtered vehicular lane line, attribute information of the vehicular lane line; and determining map data of the vehicular lane line based on the attribute information of the vehicular lane line and the positioning data during shooting of the original images. By means of the vehicular lane line data processing method, apparatus, storage medium, and device provided by the embodiments of the present disclosure, the vehicular lane line data can be efficiently and precisely determined, the labor costs in high-precision map production is greatly reduced, and the mass production of high-precision maps can be achieved.