Lane line detection method based on model fitting
A lane line detection and model fitting technology, applied in the field of computer vision, can solve problems such as line estimation errors, failure to find lane line structures, and inapplicability to unknown lane detection tasks, achieving high accuracy
Active Publication Date: 2019-06-07
QUANZHOU INST OF EQUIP MFG
View PDF4 Cites 8 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
First of all, the RANSAC algorithm is mainly proposed for the existence of only one straight line in the image. When the image data includes multiple straight lines (lane lines), it needs to be completed by the method of "fitting-and-removing" mechanism (that is, it needs to fit After a straight line, remove its corresponding interior point and continue to fit the next straight line), however, when a straight line is estimated inaccurately, it will lead to incorrect estimation of the remaining straight lines in the data
Secondly, it is necessary to specify the number of fitted straight lines to terminate the algorithm, so that RANSAC cannot be applied to the detection task of the unknown number of lane lines.
Finally, the RANSAC algorithm needs to find the interior points belonging to the line by setting the threshold in each iteration of the straight line fitting. If the threshold is too low or too high, the number of interior points will change, which will make RANSAC unable to find a reasonable lane line structure.
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0044] Such as figure 1 As shown, the lane line detection method of the embodiment of the present invention is as follows:
[0045] A lane line detection method based on model fitting includes the following steps:
[0046] Step 10. Perform a weighted average of the values of the R, G, and B channels of the color driving image to obtain a grayscale image;
[0047] Step 20: Perform Gaussian filtering on the grayscale image;
[0048] Step 30: Use Canny algorithm to obtain edge information on the grayscale image after Gaussian filtering, select a designated polygon area as a region of interest for lane line detection, and extract edge information within the region of interest;
[0049] Step 40: Establish model hypotheses based on the edge information of step 30, use least squares method for parameter calculation, assign corresponding weights to each of the model hypotheses, and remove low-weight model hypotheses by calculating information entropy and adaptive threshold;
[0050] Step 50: ...
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
Login to View More Abstract
The invention provides a lane line detection method based on model fitting. The lane line detection method comprises the following steps: carrying out weighted average on values of R, G and B channelsof a color driving image to obtain a grayscale image; carrying out gaussian filtering on the grayscale image; adopting a Canny algorithm to obtain edge information points of the grayscale image afterGaussian filtering, and extracting edge information in a region of interest; establishing a model hypothesis according to the edge information, endowing each model hypothesis with a corresponding weight, and removing the model hypothesis with a low weight by calculating an information entropy and an adaptive threshold; adopting a clustering algorithm to extract a model hypothesis related to the lane line; and fitting the parameters assumed by the model into straight lines, merging the related straight lines belonging to the same lane line, and displaying the merged straight lines in the colordriving image in an overlapping manner. The lane line detection method based on model fitting can be applied to early warning during lane departure in an automobile safety auxiliary driving system and a vehicle-mounted automobile data recorder.
Description
Technical field [0001] The invention belongs to the field of computer vision, and specifically relates to a lane line detection method based on model fitting. Background technique [0002] Lane detection technology is one of the key technologies of auto-driving and safety-assisted driving systems. Accurate and fast lane line detection is carried out through the visual method of the on-board camera, which can provide early warning for the driving vehicle when the lane deviates. A common visual method is to threshold the image first, then use edge detection algorithms to extract the contours of lane lines, and finally use random sampling agreement (RANSAC) and other algorithms to detect lane lines. [0003] The inventor found that it has certain limitations when studying the traditional RANSAC-based lane line detection method. First of all, the RANSAC algorithm is mainly proposed for the existence of only one straight line in the image. When multiple straight lines (lane lines) are ...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/46G06K9/62
Inventor 李琦铭李俊肖自能喻雷平
Owner QUANZHOU INST OF EQUIP MFG