Unlock instant, AI-driven research and patent intelligence for your innovation.

Lane line detection method based on line segment intersection-to-union ratio loss function

A technology of lane line detection and loss function, which is applied in the field of computer vision lane line detection, can solve problems such as unsuitable lane lines, achieve the effect of improving detection accuracy and overall detection accuracy

Active Publication Date: 2022-05-10
HANGZHOU FABU TECH CO LTD
View PDF10 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this type of loss function is based on a rectangular frame and is not applicable to lane lines

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
  • Lane line detection method based on line segment intersection-to-union ratio loss function
  • Lane line detection method based on line segment intersection-to-union ratio loss function
  • Lane line detection method based on line segment intersection-to-union ratio loss function

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] Below in conjunction with accompanying drawing and embodiment the present invention is described in further detail,

[0041] Such as figure 1 As shown, the embodiment and the process thereof implemented according to the complete method of the content of the present invention include the following steps:

[0042] (1) Obtain real road scene pictures, mark the actual position of the lane line as the target, and form a training data set;

[0043] In step (1), the actual position of the lane line is expressed as:

[0044]

[0045]

[0046] Among them, G is the point set of the actual position of the lane line, N is the total number of point sets, Indicates the abscissa of the actual position of the i-th lane line, is the ordinate of the actual position of the i-th lane line, H indicates the height of the picture, i indicates the ordinal number of the actual position of the lane line, and g indicates the actual position. .

[0047] (2) Input the training data se...

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 lane line detection method based on a line segment intersection-to-union ratio loss function. Acquiring a real road scene picture, marking the actual position of a lane line, and forming a training data set; inputting a detection model to obtain a lane line prediction position; constructing a line segment intersection-union ratio loss function, and calculating the classification loss of the lane lines and the regression loss of the lane lines; optimizing a loss function for training; and selecting a real-time to-be-detected road scene picture, inputting the real-time to-be-detected road scene picture into the trained detection model, and processing to obtain a lane line prediction position. The lane line is regarded as a whole for regression prediction, the loss function is simple and effective, the overall detection precision is improved, and the method has high superiority.

Description

technical field [0001] The invention belongs to a lane line image processing method in the field of computer vision lane line detection, and in particular relates to a lane line detection method based on a line segment intersection ratio loss function. Background technique [0002] Lane line detection is an important task in the field of computer vision. It is a field of mutual promotion and development with deep learning. It can be applied to automatic driving or assisted driving to provide it with road lane line information, thereby helping intelligent vehicles to better locate the vehicle position. [0003] In lane line detection, it is necessary to output the position information of the lane line in the picture. In general, to measure the similarity between two lane lines, l n loss function. But this loss function treats lane lines as multiple independent point sets, and each point is regressed separately. This assumption is too simplistic and does not consider the l...

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): G06V20/58G06V10/774G06V10/82G06K9/62G06N3/04
CPCG06N3/045G06F18/214Y02T10/40
Inventor 郑途黄亦非刘洋唐文剑杨政何晓飞
Owner HANGZHOU FABU TECH CO LTD