CN114998855A - Road edge line generation method and device, storage medium and computer equipment - Google Patents
Road edge line generation method and device, storage medium and computer equipment Download PDFInfo
- Publication number
- CN114998855A CN114998855A CN202210601094.2A CN202210601094A CN114998855A CN 114998855 A CN114998855 A CN 114998855A CN 202210601094 A CN202210601094 A CN 202210601094A CN 114998855 A CN114998855 A CN 114998855A
- Authority
- CN
- China
- Prior art keywords
- edge
- interval
- determining
- road
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30241—Trajectory
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30248—Vehicle exterior or interior
- G06T2207/30252—Vehicle exterior; Vicinity of vehicle
- G06T2207/30256—Lane; Road marking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The application discloses a road edge line generation method, a road edge line generation device, a storage medium and computer equipment, wherein the method comprises the following steps: acquiring a first acquisition image; acquiring the number of a plurality of edge areas; determining a plurality of target edge regions in the plurality of edge regions according to the number and the number threshold; acquiring running track information of a vehicle; determining a first interval and a second interval in the first collected image according to the driving track information; determining a first optimal edge area in a first interval and a second optimal edge area in a second interval according to the target edge areas and a preset ratio threshold; calculating a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area; and generating a target edge line of the road according to the plurality of first edge lines, the plurality of second edge lines and a first preset condition. The accuracy of extracting the road edge lines of the irregular road is effectively improved.
Description
Technical Field
The present application relates to the field of road boundary detection technologies, and in particular, to a method and an apparatus for generating a road edge line, a storage medium, and a computer device.
Background
In the related art, the algorithm for detecting the edge of the regular road is mainly processed by means of the obvious structure and linear characteristics in the road, such as lane lines and other information. However, the road surface in the mining area environment is an earth road or a sandstone soil block road surface, such a road surface has no uniform standard, and an irregular retaining wall and a cliff fault formed after mining exist at the road edge, so that the accuracy of the detected road edge result is low when a regular road edge detection algorithm is used for detecting the complex irregular road environment in the mining area.
Disclosure of Invention
In view of the above, the present application provides a method, an apparatus, a storage medium, and a computer device for generating a road edge line.
According to an aspect of the present application, there is provided a method of generating a road edge line, including:
responding to a generation request of a road edge line, and acquiring a first acquisition image;
determining at least one edge area of the road according to the preset example and the first collected image;
acquiring the number of a plurality of edge areas under the condition that at least one edge area is a plurality of edge areas;
determining a plurality of target edge regions within the plurality of edge regions according to the number and the number threshold;
acquiring running track information of a vehicle;
determining a first interval and a second interval in the first collected image according to the driving track information;
determining a first optimal edge area in a first interval and a second optimal edge area in a second interval according to the target edge areas and a preset ratio threshold;
calculating a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area;
and generating a target edge line of the road according to the plurality of first edge lines, the plurality of second edge lines and a first preset condition.
Optionally, the step of determining at least one edge area of the road according to the preset example and the first captured image specifically includes:
acquiring a second acquired image;
generating a training data set according to a preset example and the second collected image;
training the example segmentation network according to a training data set to obtain a pre-training weight;
reasoning is carried out on the first collected image according to the pre-training weight to generate at least one edge area;
wherein the preset examples include at least one of: retaining wall, polygonal soil heap and cliff breaking.
Optionally, the step of determining a plurality of target edge regions in the plurality of edge regions according to the number and the number threshold specifically includes:
if the number is less than or equal to the number threshold, determining the plurality of edge areas as a plurality of target edge areas;
if the number is larger than the number threshold, acquiring a plurality of width values of a plurality of edge regions, wherein the width values are numerical values of the overlapping positions of the plurality of edge regions and the first acquired image boundary;
sequencing the edge areas according to the sequence of the width values from large to small;
and determining a plurality of target edge regions in the sorted plurality of edge regions according to the number, the number threshold and the plurality of width values.
Optionally, the step of determining a first interval and a second interval in the first captured image according to the driving track information specifically includes:
performing matrix conversion on the running track information to generate first pixel point information of the vehicle on a road;
determining a road center line of the first collected image according to the first pixel point information;
and dividing the first collected image into a first section and a second section according to the center line of the road.
Optionally, the step of determining a first optimal edge area in a first interval and a second optimal edge area in a second interval according to the plurality of target edge areas and a preset ratio threshold specifically includes:
acquiring second pixel point information of each target edge area;
calculating a first ratio of each target edge region in a first interval and a second ratio of each target edge region in a second interval according to the second pixel point information;
if the first proportion value is larger than a preset proportion threshold value, determining that a target edge area corresponding to the first proportion value is located in a first interval;
if the second ratio value is larger than the preset ratio threshold value, determining that the target edge area corresponding to the second ratio value is located in a second interval;
acquiring the number of target edge areas in a first interval;
under the condition that a plurality of target edge areas exist in the first interval, acquiring the pixel position of each target edge area on a first preset boundary line of the first collected image;
sequencing a plurality of target edge areas in a first interval according to the sequence of pixel positions from low to high;
determining a first optimal edge area in the first interval according to a second preset condition;
acquiring the number of target edge areas in a second interval;
under the condition that a plurality of target edge areas exist in the second interval, acquiring the pixel position of each target edge area on a second preset boundary line of the first collected image;
sequencing the plurality of target edge areas in the second interval according to the sequence of the pixel positions from low to high;
and determining a second optimal edge area in a second interval according to a second preset condition.
Optionally, the step of calculating a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area specifically includes:
respectively extracting third pixel point information of the first optimal edge area and fourth pixel point information of the second optimal edge area by using a canny edge detection algorithm to generate a plurality of first edge pixel point sets and a plurality of second edge pixel point sets;
respectively performing second-order polynomial fitting on each first edge pixel point set and each second edge pixel point set by using a least square method to generate an edge equation y which is ax + b, wherein x is the abscissa of each pixel point, y is the ordinate of each pixel point, a is the slope, and b is the intercept;
determining a plurality of first coordinate position points of the first optimal edge area and a plurality of second coordinate position points of the second optimal edge area according to the edge equation and a third preset condition;
determining a plurality of corresponding first edge lines according to the plurality of first coordinate position points;
and determining a plurality of corresponding second edge lines according to the plurality of second coordinate position points.
Optionally, the method further comprises: and determining invalid pixel points according to the edge equation and the preset distance, and deleting the invalid pixel points in the optimal edge region.
According to another aspect of the present application, there is provided a road edge line generation apparatus including:
the first acquisition module is used for responding to a generation request of a road edge line and acquiring a first acquisition image;
the first determining module is used for determining at least one edge area of the road according to a preset example and the first collected image;
the second acquisition module is used for acquiring the number of the plurality of edge areas under the condition that at least one edge area is a plurality of edge areas;
a second determining module, configured to determine a plurality of target edge regions in the plurality of edge regions according to the number and the number threshold;
the third acquisition module is used for acquiring the running track information of the vehicle;
the fourth determining module is used for determining a first interval and a second interval in the first collected image according to the running track information;
a fifth determining module, configured to determine a first optimal edge area in the first interval and a second optimal edge area in the second interval according to the multiple target edge areas and a preset occupancy threshold;
the calculating module is used for calculating a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area;
and the generating module is used for generating the target edge line of the road according to the plurality of first edge lines, the plurality of second edge lines and the first preset condition.
Optionally, the first determining module is specifically configured to:
acquiring a second acquired image;
generating a training data set according to a preset example and the second collected image;
training the example segmentation network according to a training data set to obtain a pre-training weight;
reasoning is carried out on the first collected image according to the pre-training weight to generate at least one edge area;
wherein the preset examples include at least one of: retaining wall, polygonal soil pile and cliff.
Optionally, the second determining module is specifically configured to:
if the number is less than or equal to the number threshold, determining the plurality of edge areas as a plurality of target edge areas;
if the number is larger than the number threshold, acquiring a plurality of width values of a plurality of edge regions, wherein the width values are numerical values of the overlapping positions of the plurality of edge regions and the first acquired image boundary;
sequencing the edge areas according to the sequence of the width values from large to small;
and determining a plurality of target edge regions in the sorted plurality of edge regions according to the number, the number threshold and the plurality of width values.
Optionally, the third determining module is specifically configured to:
performing matrix conversion on the running track information to generate first pixel point information of the vehicle on a road;
determining a road center line of the first collected image according to the first pixel point information;
and dividing the first collected image into a first section and a second section according to the center line of the road.
Optionally, the fourth determining module is specifically configured to:
acquiring second pixel point information of each target edge area;
calculating a first ratio of each target edge region in a first interval and a second ratio of each target edge region in a second interval according to the second pixel point information;
if the first ratio value is larger than a preset ratio threshold value, determining that a target edge area corresponding to the first ratio value is located in a first interval;
if the second ratio value is larger than the preset ratio threshold value, determining that the target edge area corresponding to the second ratio value is located in a second interval;
acquiring the number of target edge areas in a first interval;
under the condition that a plurality of target edge areas exist in the first interval, acquiring the pixel position of each target edge area on a first preset boundary line of the first collected image;
sequencing a plurality of target edge areas in a first interval according to the sequence of pixel positions from low to high;
determining a first optimal edge area in the first interval according to a second preset condition;
acquiring the number of target edge areas in a second interval;
under the condition that a plurality of target edge areas exist in the second interval, acquiring the pixel position of each target edge area on a second preset boundary line of the first collected image;
sequencing the plurality of target edge areas in the second interval according to the sequence of the pixel positions from low to high;
and determining a second optimal edge area in a second interval according to a second preset condition.
Optionally, the calculation module is specifically configured to:
respectively extracting third pixel point information of the first optimal edge area and fourth pixel point information of the second optimal edge area by using a canny edge detection algorithm to generate a plurality of first edge pixel point sets and a plurality of second edge pixel point sets;
respectively performing second-order polynomial fitting on each first edge pixel point set and each second edge pixel point set by using a least square method to generate an edge equation y which is ax + b, wherein x is the abscissa of each pixel point, y is the ordinate of each pixel point, a is the slope, and b is the intercept;
determining a plurality of first coordinate position points of the first optimal edge area and a plurality of second coordinate position points of the second optimal edge area according to an edge equation and a third preset condition;
determining a plurality of corresponding first edge lines according to the plurality of first coordinate position points;
and determining a plurality of corresponding second edge lines according to the plurality of second coordinate position points.
Optionally, the apparatus for generating a road edge line further includes:
and the sixth determining module is used for determining invalid pixel points according to the edge equation and the preset distance and deleting the invalid pixel points in the optimal edge region.
By means of the technical scheme, the method for generating the road edge line detects the irregular retaining wall and the cliff example in the image by using the example segmentation algorithm to serve as the irregular road edge example, the irregular road example is classified and processed, and then the irregular road edge detection is carried out, so that the extracted road edge line is more suitable for the irregular road working condition of a mining area, and the accuracy and the precision of the irregular road edge line are effectively improved.
The foregoing description is only an overview of the technical solutions of the present application, and the present application can be implemented according to the content of the description in order to make the technical means of the present application more clearly understood, and the following detailed description of the present application is given in order to make the above and other objects, features, and advantages of the present application more clearly understandable.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
fig. 1 is a schematic flowchart illustrating a method for generating a road edge line according to an embodiment of the present application;
FIG. 2 is a schematic diagram illustrating a plurality of target edge regions of a first captured image provided by an embodiment of the present application;
fig. 3 is a schematic diagram illustrating a pixel point on a target edge line according to an embodiment of the present disclosure;
FIG. 4 is a schematic diagram illustrating a plurality of edge lines of an optimal edge region of a first captured image according to an embodiment of the present disclosure;
fig. 5 is a schematic structural diagram illustrating a road edge line generation device according to an embodiment of the present application.
Detailed Description
The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
In the present embodiment, there is provided a method for generating a road edge line, as shown in fig. 1, the method including:
104, determining at least one edge area of the road according to a preset example and the first acquired image;
step 110, determining a plurality of target edge regions in the plurality of edge regions according to the number and the number threshold;
step 118, calculating a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area;
and step 120, determining a target edge line of the road according to the plurality of first edge lines, the plurality of second edge lines and the first preset condition.
The embodiment of the application provides a method for generating a road edge line, and particularly, in the process of automatic driving of a vehicle in a mining area, a first collected image is obtained in response to a request for generating the road edge line, wherein the first collected image refers to a received real-time road image collected by a vehicle sensor system located at the front end of the vehicle. And then, determining at least one edge area in the first acquired image according to preset examples and the acquired real-time road image, wherein the preset examples refer to road edge retaining walls of irregular roads in the mining area, irregular soil piles, cliffs generated by mining in the mining area and other road edge area examples needing to be identified. Further, according to three examples, namely a road edge retaining wall, an irregular soil pile and a cliff, reasoning is carried out on road images acquired in the real-time driving process of the vehicle by using an example segmentation algorithm, and at least one edge area of the irregular road in the first acquired image is identified.
Further, since there are usually more irregular road edge areas in the driving process of the vehicle in the mining area, when it is determined that the edge areas of the irregular road are multiple edge areas, the multiple edge areas need to be processed in advance, so as to determine the edge area closest to the driving track of the vehicle on one side in the multiple edge areas. Specifically, the number of the plurality of edge regions is obtained, and a plurality of target edge regions meeting the number threshold range are screened out from the plurality of edge regions according to the number of the plurality of edge regions and the number threshold. And then, receiving the driving track information of the vehicle sent by the vehicle controller, determining the road center line of the first collected image according to the driving track information of the vehicle, and further dividing the first collected image into a first section and a second section according to the road center line, wherein the first section refers to the section on the left side of the road center line in the collected image, and the second section refers to the section on the right side of the road center line in the collected image. After the left and right intervals of the first collected image are determined, the occupation ratios of a plurality of target edge areas in a first interval and a second interval are respectively calculated, and according to the occupation ratios of the plurality of target edge areas in the first interval and the second interval and a preset occupation ratio threshold, a first optimal edge area in the first interval and a second optimal edge area in the second interval are determined, wherein the first optimal edge area refers to the edge area which is located in the left interval in the first collected image and is closest to the vehicle running track, and the second optimal edge area refers to the edge area which is located in the right interval in the first collected image and is closest to the vehicle running track.
It can be understood that, when the edge area of the irregular road is judged to be single, the edge area is the optimal edge area in the first captured image.
Further, a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area are respectively calculated. And then screening out edge lines meeting a first preset condition in the plurality of first edge lines, namely target edge lines in a first interval, and screening out edge lines meeting the first preset condition in the plurality of second edge lines, namely target edge lines in a second interval, wherein the first preset condition is the edge line with the pixel point closest to the lower edge of the image.
By applying the technical scheme of the embodiment, the retaining wall, the irregular soil heap and the cliff are regarded as the irregular road edge examples to be identified based on the specific mining area environment, the real-time road image is processed to determine the edge area of the irregular road, and then the irregular road edge lines are extracted from the determined edge area of the irregular road, so that the accuracy and the precision of extracting the irregular road edge lines are effectively improved.
Optionally, the number threshold may be set to 4, and by presetting the number threshold, the processing time for extracting the irregular road edge line is effectively reduced, and the occupancy rate of the computing resource is reduced.
In this embodiment of the present application, optionally, step 104 may specifically include: acquiring a second acquired image; generating a training data set according to a preset example and the second acquired image; training the example segmentation network according to a training data set to obtain a pre-training weight; reasoning is carried out on the first collected image according to the pre-training weight to generate at least one edge area; wherein the preset examples include at least one of: retaining wall, polygonal soil heap and cliff breaking.
In this embodiment, a second collected image is obtained, where the second collected image refers to a collected mining area-related road condition image, and specifically may be a historical mining area road condition image. And after obtaining the relevant road condition images of the mining area, carrying out irregular polygon labeling on the second collected image according to three preset examples of a road edge retaining wall, an irregular soil heap and a cliff to generate a training data set. Further, training the example segmentation network by using a training data set to obtain pre-training weights for irregular area example segmentation in corresponding scenes. And then, reasoning the first collected image by using the pre-training weight to obtain at least one edge area, namely the edge area of the irregular road.
By the mode, the road edge retaining wall, the irregular soil heap and the cliff serve as preset examples, the mine area irregular road edge area is detected by adopting the example-based segmentation network, and compared with a regular road detection method for detecting the edge by adopting regular lane lines in the prior art, the method has the advantages that the irregular retaining wall and the cliff in the image are detected by adopting the example segmentation algorithm and serve as the irregular road edge examples, the irregular road examples are classified, the accuracy of irregular road edge area identification is improved, the determined road edge area is more suitable for the mine area irregular road, and the accuracy of follow-up irregular road edge line extraction is ensured.
In this embodiment of the present application, optionally, step 110 may specifically include: if the number is less than or equal to the number threshold, determining the plurality of edge areas as a plurality of target edge areas; if the number is larger than the number threshold, acquiring a plurality of width values of a plurality of edge regions, wherein the width values are numerical values of the overlapping positions of the plurality of edge regions and the first acquired image boundary; sequencing the edge areas according to the sequence of the width values from large to small; and determining a plurality of target edge regions in the sorted plurality of edge regions according to the number, the number threshold and the plurality of width values.
In this embodiment, after identifying a number of edge regions included within the first captured image, the number of edge regions is compared to a number threshold. When the number of the plurality of edge regions is smaller than the number threshold, it is indicated that the number of the edge regions in the first captured image is within a preset range, and at this time, all of the plurality of edge regions are taken as target edge regions. Further, when the number of the plurality of edge regions is greater than the number threshold, it is indicated that the number of the edge regions in the first captured image exceeds the preset range. At this time, the width value of each edge region, that is, the numerical value of the overlapping position of each edge region and the side boundary of the first acquired image, is acquired, the plurality of edge regions are sorted in the order of the width values from large to small, and the edge region with the width value arranged in front is screened out as the target edge region in the sorted plurality of edge regions according to the number threshold.
By the method, the target edge area is reserved according to the width of the edge area and the set number range, the processing time for subsequently extracting the edge line is effectively reduced, and the occupancy rate of the computing resource is reduced.
In a specific embodiment, the number threshold is set to be 4, and when the number of the plurality of edge regions is less than or equal to 4, all the plurality of edge regions are taken as target edge regions; when the number of the edge areas is more than 4, the widths of the edge areas are sequenced, and the first 4 edge areas with the widest width are reserved as target edge areas, so that the maximum number of the edge areas reserved in the acquired image is 4.
In this embodiment, optionally, step 114 may specifically include: performing matrix conversion on the running track information to generate first pixel point information of the vehicle on a road; determining a road center line of the first collected image according to the first pixel point information; and dividing the first collected image into a first section and a second section according to the center line of the road.
In this embodiment, after the driving track information of the vehicle is collected, the driving track information of the vehicle is converted into image pixel point information of the vehicle on the current road, that is, first pixel point information, by converting the driving track information into a matrix. In the automatic driving process of the vehicle, the driving track of the vehicle on the road can be regarded as the center line of the current road, namely, the first pixel point information is regarded as the center line of the road on the first collected image. Further, after the center line of the road is determined, the first collected image is divided into a first section and a second section according to the left side and the right side of the center line of the road, wherein the first section is a section on the left side of the center line of the road in the first collected image, and the second section is a section on the right side of the center line of the road in the first collected image.
By the mode, the driving path of the vehicle on the real-time collected image can be accurately positioned by utilizing the driving track of the vehicle, meanwhile, the road center line determined by the driving track information is utilized, the collected image is divided into a left section and a right section according to the road center line, and the accuracy of the detection of the edge of the irregular road in the mining area is improved by combining the driving track information of the vehicle and the collected image.
It can be understood that the driving track information of the vehicle is the current driving path information of the vehicle, and in the automatic driving process of the vehicle, a driving point is collected at a preset distance through a Global Positioning System (GPS) on the vehicle, for example, a point is collected when the vehicle drives for 60 meters, and then the driving track information of the vehicle is obtained and sent through a controller. After receiving the driving track information sent by the vehicle, converting the three-dimensional point information into two-dimensional point information through the conversion matrix to obtain image pixel point information.
In this embodiment of the application, optionally, step 116 may specifically include: acquiring second pixel point information of each target edge area; calculating a first ratio of each target edge region in a first interval and a second ratio of each target edge region in a second interval according to the second pixel point information; if the first ratio value is larger than a preset ratio threshold value, determining that at least one target edge area corresponding to the first ratio value is located in a first interval; if the second ratio value is larger than the preset ratio threshold value, determining that at least one target edge area corresponding to the second ratio value is located in a second interval; under the condition that at least one target edge area is multiple, acquiring the pixel position of each target edge area at the boundary of a first acquired image; respectively sequencing a plurality of target edge areas in a first interval and a second interval according to the sequence of pixel positions from low to high; and determining a first optimal edge area in the first interval and a second optimal edge area in the second interval according to the preset position.
In this embodiment, second pixel point information of each target edge region in the first acquired image is acquired, the number of pixel points of each edge target region is determined, and then a first ratio of each target edge region in a first interval is calculated according to the number of pixel points of each target edge region, that is, the ratio of the target edge region in a left interval of the image is calculated. Meanwhile, according to the number of the pixel points of each target edge region, calculating a second ratio of each target edge region in a second interval, namely calculating the ratio of the target edge region in the right interval of the image.
Further, after the proportion value of each target edge region in the left and right intervals is calculated respectively, each target edge region is judged to belong to the first interval or the second interval according to a preset proportion threshold value. Specifically, when the occupation ratio value of the target edge area in the first section is greater than the preset occupation ratio threshold, the target edge area is indicated as the edge area on the left side of the image, and the target edge area is divided into the first section. Further, when the occupation ratio value of the target edge area in the second interval is greater than the preset occupation ratio threshold, determining that the target edge area is the edge area on the right side of the image, and dividing the target edge area into the second interval.
Further, the number of target edge areas in the first interval and the number of target edge areas in the second interval are respectively obtained, when the number of the target edge areas in the first interval or the second interval is multiple, the pixel position of the overlapping position of each target edge area in the first interval or the second interval and a preset boundary line on the first collected image is determined, and the optimal edge areas of the first interval and the second interval are respectively determined according to the pixel positions of the multiple target edge areas.
Specifically, when a plurality of target edge regions are located in the first interval, a pixel position of each target edge region on a first preset boundary line in the first captured image is obtained, where it should be noted that the first preset boundary line is an image left boundary line. And then, determining the lowest pixel position in the pixel positions of each target edge region, sequencing the lowest pixel positions of the target edge regions from low to high on the left boundary line of the image, and screening out the target edge region meeting a second preset condition as an optimal edge region, wherein the second preset condition is the pixel position which is closest to the bottom of the image in all the pixel positions. Namely, the target edge region corresponding to the pixel position closest to the bottom of the image is reserved as the first optimal edge region of the first interval.
Further, when a plurality of target edge regions are included in the second interval, a pixel position of each target edge region on a second preset boundary line in the first captured image is obtained, where the second preset boundary line is a right boundary line of the first captured image, that is, a pixel position of each target edge region in the second interval on a right boundary line of the image is obtained. And then, determining the lowest pixel position in the pixel positions of each target edge region, sequencing the lowest pixel positions of the target edge regions from low to high on the right boundary line of the image, and reserving the target edge region corresponding to the pixel position closest to the bottom of the image as the second optimal edge region of the second interval.
By the method, the edge area closest to the bottom of the image is determined in the left and right side areas of the collected image respectively, and the determined optimal edge area is used for determining the road edge line, so that the calculation amount of subsequent edge line extraction is effectively reduced, and the calculation difficulty of the system is reduced.
In a specific embodiment, as shown in fig. 2, the first captured image is a schematic diagram of a plurality of target edge areas, in which a target edge area a is located in a first interval, and since the target edge area in the first interval is a single target edge area, the target edge area a is used as a first optimal edge area of the first interval. Further, in the figure, the target edge area B and the target edge area C are located in the second interval, and the optimal edge area of the second interval needs to be determined according to the pixel positions of the target edge area B and the target edge area C. Specifically, the acquisition target edge region B is located at the pixel positions B1, B2 on the right boundary line of the captured image, while the acquisition target edge region C is located at the pixel positions C1, C2 on the right boundary line of the captured image. The lowest pixel position B2 of the target edge region B and the lowest pixel position C2 of the target edge region C are sorted in descending order on the right boundary line of the image, and C2 is determined as the pixel position closest to the bottom of the image, so that the target edge region C is determined as the target edge region closest to the bottom of the image, and the target edge region C is determined as the second best edge region of the second section.
In this embodiment of the application, optionally, step 118 may specifically include: respectively extracting pixel point information of a first optimal edge area and pixel point information of a second optimal edge area by using a canny edge detection algorithm to generate a plurality of first edge pixel point sets and a plurality of second edge pixel point sets; respectively performing second-order polynomial fitting on each first edge pixel point set and each second edge pixel point set by using a least square method to generate an edge equation y which is ax + b, wherein x and y are respectively the horizontal and vertical coordinates of each pixel point, a is a slope, and b is an intercept; determining a plurality of first coordinate position points of the first optimal edge area and a plurality of second coordinate position points of the second optimal edge area according to an edge equation and a third preset condition; determining a plurality of corresponding first edge lines according to the plurality of first coordinate position points; and determining a plurality of corresponding second edge lines according to the plurality of second coordinate position points.
In this embodiment, edge pixel point information of a plurality of edge lines in the left and right edge regions is extracted respectively by a canny edge detection algorithm, so as to obtain a plurality of first edge pixel point sets and a plurality of second edge pixel point sets. Then, each edge pixel point set extracted by the canny edge detection algorithm is fitted to an edge equation y of horizontal and vertical coordinate information about the pixel point by a least square method, and the position information of each pixel point can be expressed by the position information of the horizontal and vertical coordinate axes. Further, according to the edge equation, a plurality of first coordinate position points and a plurality of second coordinate position points which meet a third preset condition in the first interval and the second interval are respectively calculated, wherein the third preset condition is a longitudinal coordinate position point of a pixel at which a horizontal pixel of the edge equation is 0. That is, a plurality of vertical coordinate position points of the pixel where the left edge equation is at 0 in the horizontal pixel point are calculated, and a plurality of vertical coordinate position points of the pixel where the right edge equation is at 0 in the horizontal pixel point are calculated. And then determining a plurality of corresponding edge lines according to the plurality of calculated first coordinate position points, and determining a plurality of corresponding edge lines according to the plurality of calculated second coordinate position points.
It should be noted that the slope of the edge equation is determined as the slope requirement of the first interval being greater than 0, and the slope requirement of the second interval being less than 0.
By the method, a plurality of edge lines of the left and right intervals of the acquired image are calculated by using a canny edge detection algorithm and combining a least square method, so that the accuracy of extracting the edge lines of the image edge areas is ensured.
In a specific embodiment, as shown in fig. 4, a plurality of edge lines of the optimal edge region of the first captured image are shown, specifically, after the optimal edge regions of the left and right two regions of the captured image are determined respectively, calculating longitudinal coordinate position points p1 and p2 of a first optimal edge area of the first interval at a transverse pixel of 0 by using a canny edge detection algorithm in combination with a least square method, and the vertical coordinate position points p3 and p4 of the second optimal edge region of the second section at the horizontal pixel of 0, thereafter, the corresponding edge line 1 is determined according to the determined first coordinate position point p1, the corresponding edge line 2 is determined according to the determined first coordinate position point p2, the corresponding edge line 3 is determined according to the determined second coordinate position point p3, and the corresponding edge line 4 is determined according to the determined second coordinate position point p 4. Further, after determining the edge line 1, the edge line 2, the edge line 3 and the edge line 4, sorting the edge lines reserved in the left and right sections of the acquired image according to the sequence from low to high of the edge points located in the boundary line of the image, and reserving the edge lines of which the edge points are closest to the lower edge of the image as target edge lines on the left and right sides of the acquired image, namely the edge line 2 and the edge line 4 in the image.
In the embodiment of the present application, optionally, the method further includes: and determining invalid pixel points according to the edge equation and the preset distance, and deleting the invalid pixel points in the optimal edge region.
In this embodiment, for a single road in the mining area environment, the directions of the roads should be kept consistent in a state of sufficient length, and therefore, after fitting the edge equations on the left and right sides, the distances from different points to the line are calculated according to the edge equations, if the calculated distance is greater than the preset distance, it is indicated that a road with abrupt transition exists, the point is determined to be an invalid pixel point, and the determined invalid pixel point is deleted in the road edge line, so as to ensure the accuracy of the finally generated road edge line.
It will be appreciated that the preset distance may be an average of the distances of the different points to the line.
Optionally, as shown in fig. 3, the schematic diagram of the pixel points on the target edge line is shown, after the target edge line is determined, the discrete points and the missing points in the road edge line are determined according to the discrete degree from the road, and then the discrete points are deleted and/or the missing points are supplemented, so as to ensure the continuity of the finally extracted road edge and improve the accuracy of the irregular road edge line.
An algorithm which is based on example segmentation and is suitable for extracting the irregular edge lines of the roads in the mining area is provided, the irregular road edges which can guide the automatic driving vehicles to run are obtained, and the safety and the stability of automatic driving are improved.
In one embodiment of the application, a method for generating road edge lines is provided, wherein irregular retaining walls and cliff examples in related images are detected by using an example segmentation algorithm to serve as irregular road edge examples. And then, extracting the edge line of the irregular road through the steps of irregular road example classification processing, irregular road edge detection and irregular road edge optimization, and applying the edge line to automatic driving of the road in the mining area to realize the edge detection of the irregular road.
Specifically, because the road edge retaining wall and the cliff in the mining area are in irregular shapes, the region of interest can not be detected through modeling in a regular shape, and therefore an example segmentation model taking a polygon as output is selected for training to obtain the irregular polygonal region of interest. That is, in a mine area environment, a road edge retaining wall of a current lane on which a vehicle travels, a side-shaped soil heap, and a cliff due to mining of the mine area are regarded as irregular road edge area examples, i.e., preset examples, that need to be identified.
Furthermore, relevant road condition pictures, namely second collected images, are collected in the mining area, and irregular polygon labeling is carried out according to 3 preset examples, namely a road edge retaining wall, an irregular soil heap and a cliff, and is used as a training data set for example segmentation. And then, training the example segmentation network by using a training data set to obtain pre-training weights of the irregular area example segmentation in the corresponding scene. And reasoning a first acquired image acquired in real vehicle running by using the pre-training weight to obtain an irregular road edge area.
Further, after the road edge regions are obtained from a single first collected image, the number of the edge regions is calculated, and the irregular road edge regions are divided into a single edge region and a plurality of edge regions for processing.
Further, when it is determined that the edge area of the road is a plurality of edge areas, the edge area of the irregular road is usually longer in the driving process due to the processing of the plurality of edge areas, and therefore the misidentification example needs to be deleted. And then judging the position area of the edge area in the acquired image, distinguishing the left side and the right side of the vehicle, further respectively judging the position state according to the left edge and the right edge, and finally finishing that at most one example is reserved on one side.
Specifically, the number of a plurality of edge regions is obtained, and the number of regions extracted by the algorithm is greater than a number threshold, for example, when 4 regions are extracted, the width lengths of the edges of the regions are sorted, and the first 4 regions with the widest width are reserved. Through the steps, the number of the plurality of edge areas is ensured to be at most 4. Subsequently, the 4 zones are divided into left and right retaining wall portions, i.e., a first zone and a second zone. The method comprises the steps of converting driving track information of a vehicle into image pixel point information of a current lane, calculating the ratio of each edge area on the left side and the right side of a lane pixel, and dividing the edge area into a left side area when the ratio of the pixels on the left side exceeds a preset ratio threshold value, such as 75%; when the right pixel occupancy exceeds a preset occupancy threshold, for example 75%, the edge area is divided into right areas. Further, the left and right side edge areas are sorted according to the position of the edge pixel of the lowest edge pixel, and the edge areas with one side closest to the bottom of the image are respectively reserved. By this step, the optimum edge area of the one-side closest vehicle travel area can be obtained.
Further, when the edge area of the irregular road is determined to be a single edge area, the edge area can be directly used as the optimal edge area of the one-side nearest vehicle driving area.
Further, after the optimal edge area of the driving area of the nearest vehicle on one side is determined, the edge line of the edge area is extracted by using a canny edge detection algorithm according to the optimal edge areas on the left side and the right side. And then, selecting two lines closest to the vehicle direction, and performing discrete sampling to obtain the edge line of the irregular road in the vehicle driving process.
Specifically, edge pixel point sets of the left and right edge regions are respectively extracted by using a canny edge detection algorithm. Fitting each edge pixel point set extracted by the canny edge algorithm into a linear equation of horizontal and vertical coordinate information of pixel points by a least square method: and further, calculating a longitudinal coordinate position point of a pixel of the left edge equation at the position of a horizontal pixel of 0 and a longitudinal coordinate position point of a pixel of the right edge equation at the position of a horizontal pixel of 0, and reserving an edge line of the edge point closest to the lower edge of the image as a target edge line.
Furthermore, since the irregular road edge line calculated by the method has discrete points and discontinuous areas, discrete point post-processing and edge continuity optimization need to be added, the extraction accuracy of the irregular road edge line is improved, and the final road edge line is output. Specifically, discrete points at the edge of the road are searched, and the discrete points are selected and deleted according to the discrete degree of the distance road to supplement the lost points.
Further, for a single road, the direction of the road should be consistent in a state of sufficient length, and therefore, optimization processing needs to be performed on the road with abrupt transition.
The road boundary line generation method provided by the embodiment of the application realizes edge detection of irregular roads in specific mining area environments, and has environmental adaptability and stability.
Further, as a specific implementation of the method in fig. 1, an embodiment of the present application provides an apparatus for generating a road edge line, as shown in fig. 5, the apparatus includes:
the first acquisition module is used for responding to a generation request of a road edge line and acquiring a first acquisition image;
the first determining module is used for determining at least one edge area of the road according to the preset example and the first collected image;
the second acquisition module is used for acquiring the number of the plurality of edge areas under the condition that at least one edge area is a plurality of edge areas;
a second determining module, configured to determine a plurality of target edge regions in the plurality of edge regions according to the number and the number threshold;
the third acquisition module is used for acquiring the running track information of the vehicle;
the fourth determining module is used for determining a first interval and a second interval in the first collected image according to the running track information;
a fifth determining module, configured to determine, according to the multiple target edge areas and a preset ratio threshold, a first optimal edge area in the first interval and a second optimal edge area in the second interval;
a calculation module for calculating a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area;
and the generating module is used for generating the target edge line of the road according to the plurality of first edge lines, the plurality of second edge lines and the first preset condition.
Optionally, the first determining module is specifically configured to:
acquiring a second acquired image;
generating a training data set according to a preset example and the second collected image;
training the example segmentation network according to a training data set to obtain a pre-training weight;
reasoning is carried out on the first collected image according to the pre-training weight to generate at least one edge area;
wherein the preset examples include at least one of: retaining wall, polygonal soil pile and cliff.
Optionally, the second determining module is specifically configured to:
if the number is less than or equal to the number threshold, determining the plurality of edge areas as a plurality of target edge areas;
if the number is larger than the number threshold, acquiring a plurality of width values of a plurality of edge regions, wherein the width values are numerical values of the overlapping positions of the plurality of edge regions and the first acquired image boundary;
sequencing the edge areas according to the sequence of the width values from large to small;
and determining a plurality of target edge regions in the sorted plurality of edge regions according to the number, the number threshold and the plurality of width values.
Optionally, the third determining module is specifically configured to:
performing matrix conversion on the running track information to generate first pixel point information of the vehicle on the road;
determining a road center line of the first collected image according to the first pixel point information;
and dividing the first collected image into a first section and a second section according to the center line of the road.
Optionally, the fourth determining module is specifically configured to:
acquiring second pixel point information of each target edge area;
calculating a first ratio of each target edge region in a first interval and a second ratio of each target edge region in a second interval according to the second pixel point information;
if the first ratio value is larger than a preset ratio threshold value, determining that a target edge area corresponding to the first ratio value is located in a first interval;
if the second proportion value is larger than the preset proportion threshold value, determining that the target edge area corresponding to the second proportion value is located in a second interval;
acquiring the number of target edge areas in a first interval;
under the condition that a plurality of target edge areas exist in the first interval, acquiring the pixel position of each target edge area on a first preset boundary line of the first collected image;
sequencing a plurality of target edge areas in a first interval according to the sequence of pixel positions from low to high;
determining a first optimal edge area in the first interval according to a second preset condition;
acquiring the number of target edge areas in a second interval;
under the condition that a plurality of target edge areas exist in the second interval, acquiring the pixel position of each target edge area on a second preset boundary line of the first collected image;
sequencing the plurality of target edge areas in the second interval according to the sequence of the pixel positions from low to high;
and determining a second optimal edge area in a second interval according to a second preset condition.
Optionally, the calculation module is specifically configured to:
respectively extracting third pixel point information of the first optimal edge area and fourth pixel point information of the second optimal edge area by using a canny edge detection algorithm to generate a plurality of first edge pixel point sets and a plurality of second edge pixel point sets;
respectively performing second-order polynomial fitting on each first edge pixel point set and each second edge pixel point set by using a least square method to generate an edge equation y which is ax + b, wherein x is the abscissa of each pixel point, y is the ordinate of each pixel point, a is the slope, and b is the intercept;
determining a plurality of first coordinate position points of the first optimal edge area and a plurality of second coordinate position points of the second optimal edge area according to an edge equation and a third preset condition;
determining a plurality of corresponding first edge lines according to the plurality of first coordinate position points;
and determining a plurality of corresponding second edge lines according to the plurality of second coordinate position points.
Optionally, the apparatus for generating a road edge line further includes:
and the sixth determining module is used for determining invalid pixel points according to the edge equation and the preset distance and deleting the invalid pixel points in the optimal edge region.
It should be noted that other corresponding descriptions of the functional units related to the apparatus for generating a road edge line provided in the embodiment of the present application may refer to the corresponding descriptions in the method in fig. 1, and are not described herein again.
Based on the method shown in fig. 1, correspondingly, the present application further provides a storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the method for generating a road edge line shown in fig. 1.
Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.), and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the implementation scenarios of the present application.
Based on the method shown in fig. 1 and the virtual device embodiment shown in fig. 5, in order to achieve the above object, the present application further provides a computer device, which may specifically be a personal computer, a server, a network device, and the like, where the computer device includes a storage medium and a processor; a storage medium for storing a computer program; a processor for executing a computer program to implement the above-described method of generating a road edge line as shown in fig. 1.
Optionally, the computer device may also include a user interface, a network interface, a camera, Radio Frequency (RF) circuitry, sensors, audio circuitry, a WI-FI module, and so forth. The user interface may include a Display screen (Display), an input unit such as a keypad (Keyboard), etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (e.g., a bluetooth interface, WI-FI interface), etc.
It will be appreciated by those skilled in the art that the present embodiment provides a computer device architecture that is not limiting of the computer device, and that may include more or fewer components, or some components in combination, or a different arrangement of components.
The storage medium may further include an operating system and a network communication module. An operating system is a program that manages and maintains the hardware and software resources of a computer device, supporting the operation of information handling programs, as well as other software and/or programs. The network communication module is used for realizing communication among components in the storage medium and other hardware and software in the entity device.
Through the description of the above embodiment, those skilled in the art can clearly understand that the present application can be implemented by software plus a necessary general hardware platform, and can also be implemented by hardware by using an example segmentation algorithm to detect an irregular retaining wall and a cliff example in an image as an irregular road edge example, classify the irregular road example, and further perform irregular road edge detection, so that the extracted road edge line is more suitable for the irregular road condition of a mining area, and the accuracy and precision of the irregular road edge line are effectively improved.
Those skilled in the art will appreciate that the figures are merely schematic representations of one preferred implementation scenario and that the blocks or flow diagrams in the figures are not necessarily required to practice the present application. Those skilled in the art will appreciate that the modules in the devices in the implementation scenario may be distributed in the devices in the implementation scenario according to the description of the implementation scenario, or may be located in one or more devices different from the present implementation scenario with corresponding changes. The modules of the implementation scenario may be combined into one module, or may be further split into multiple sub-modules.
The above application serial numbers are for description purposes only and do not represent the superiority or inferiority of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of the present application, but the present application is not limited thereto, and any variations that can be made by those skilled in the art are intended to fall within the scope of the present application.
Claims (10)
1. A method for generating a road edge line, comprising:
responding to a generation request of a road edge line, and acquiring a first acquisition image;
determining at least one edge area of the road according to a preset example and the first acquired image;
acquiring the number of a plurality of edge areas under the condition that the at least one edge area is a plurality of edge areas;
determining a plurality of target edge regions within the plurality of edge regions according to the number and a number threshold;
acquiring running track information of a vehicle;
determining a first interval and a second interval in the first collected image according to the running track information;
determining a first optimal edge area in the first interval and a second optimal edge area in the second interval according to the target edge areas and a preset ratio threshold;
calculating a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area;
and generating the target edge lines of the road according to the plurality of first edge lines, the plurality of second edge lines and a first preset condition.
2. The method according to claim 1, characterized in that said step of determining at least one edge region of the road from a preset instance and said first captured image comprises in particular:
acquiring a second acquired image;
generating a training data set according to the preset example and the second collected image;
training an example segmentation network according to the training data set to obtain a pre-training weight;
reasoning is carried out on the first collected image according to the pre-training weight to generate the at least one edge area;
wherein the preset instance comprises at least one of: retaining wall, polygonal soil pile and cliff.
3. The method according to claim 1, wherein the step of determining a plurality of target edge regions within the plurality of edge regions according to the number and the number threshold specifically comprises:
if the number is less than or equal to the number threshold, determining the plurality of edge areas as the plurality of target edge areas;
if the number is larger than the number threshold, acquiring a plurality of width values of the plurality of edge regions, wherein the width values are numerical values of the coincidence positions of the plurality of edge regions and the first acquired image boundary;
sequencing the edge areas according to the sequence of the width values from large to small;
determining the plurality of target edge regions within the sorted plurality of edge regions according to the number, the number threshold, and the plurality of width values.
4. The method according to claim 1, wherein the step of determining the first and second intervals in the first captured image according to the driving trajectory information includes:
performing matrix conversion on the driving track information to generate first pixel point information of the vehicle on the road;
determining a road center line of the first collected image according to the first pixel point information;
and dividing the first collected image into the first section and the second section according to the road center line.
5. The method according to claim 1, wherein the step of determining a first optimal edge region in a first interval and a second optimal edge region in a second interval according to a plurality of target edge regions and a preset duty ratio threshold specifically comprises:
acquiring second pixel point information of each target edge area;
calculating a first ratio of each target edge region in the first interval and a second ratio of each target edge region in the second interval according to the second pixel point information;
if the first occupation ratio value is larger than the preset occupation ratio threshold, determining that the target edge area corresponding to the first occupation ratio value is located in the first interval;
if the second ratio value is larger than the preset ratio threshold, determining that the target edge area corresponding to the second ratio value is located in the second interval;
acquiring the number of target edge areas in the first interval;
under the condition that a plurality of target edge areas exist in the first interval, acquiring the pixel position of each target edge area on a first preset boundary line of the first collected image;
sequencing the plurality of target edge areas in the first interval according to the sequence of the pixel positions from low to high;
determining the first optimal edge area in the first interval according to a second preset condition;
acquiring the number of target edge areas in the second interval;
under the condition that a plurality of target edge areas exist in the second interval, acquiring the pixel position of each target edge area on a second preset boundary line of the first collected image;
sequencing the plurality of target edge areas in the second interval according to the sequence of the pixel positions from low to high;
and determining the second optimal edge area in the second interval according to the second preset condition.
6. The method according to claim 1, wherein the step of calculating a plurality of first edge lines of the first optimal edge region and a plurality of second edge lines of the second optimal edge region comprises:
respectively extracting third pixel point information of the first optimal edge area and fourth pixel point information of the second optimal edge area by using a canny edge detection algorithm to generate a plurality of first edge pixel point sets and a plurality of second edge pixel point sets;
respectively performing second-order polynomial fitting on each first edge pixel point set and each second edge pixel point set by using a least square method to generate an edge equation y which is ax + b, wherein x is the abscissa of each pixel point, y is the ordinate of each pixel point, a is the slope, and b is the intercept;
determining a plurality of first coordinate position points of the first optimal edge area and a plurality of second coordinate position points of the second optimal edge area according to the edge equation and a third preset condition;
determining a plurality of first edge lines corresponding to the first coordinate position points according to the first coordinate position points;
and determining the plurality of second edge lines corresponding to the plurality of second coordinate position points according to the plurality of second coordinate position points.
7. The method according to any one of claims 1 to 6, further comprising:
and determining invalid pixel points according to the edge equation and a preset distance, and deleting the invalid pixel points in the optimal edge region.
8. An apparatus for generating a road edge line, comprising:
the first acquisition module is used for responding to a generation request of a road edge line and acquiring a first acquisition image;
the first determining module is used for determining at least one edge area of the road according to a preset example and the first collected image;
a second obtaining module, configured to obtain the number of the multiple edge areas when the at least one edge area is multiple;
a second determining module, configured to determine, according to the number and the number threshold, a plurality of target edge regions in the plurality of edge regions;
the third acquisition module is used for acquiring the running track information of the vehicle;
the fourth determining module is used for determining a first interval and a second interval in the first collected image according to the running track information;
a fifth determining module, configured to determine, according to the multiple target edge areas and a preset duty threshold, a first optimal edge area in the first interval and a second optimal edge area in the second interval;
a calculation module, configured to calculate a plurality of first edge lines of the first optimal edge area and a plurality of second edge lines of the second optimal edge area;
and the generating module is used for generating the target edge line of the road according to the plurality of first edge lines, the plurality of second edge lines and a first preset condition.
9. A storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method of any of claims 1 to 7.
10. A computer device comprising a storage medium, a processor and a computer program stored on the storage medium and executable on the processor, characterized in that the processor implements the method of any one of claims 1 to 7 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210601094.2A CN114998855A (en) | 2022-05-30 | 2022-05-30 | Road edge line generation method and device, storage medium and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210601094.2A CN114998855A (en) | 2022-05-30 | 2022-05-30 | Road edge line generation method and device, storage medium and computer equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114998855A true CN114998855A (en) | 2022-09-02 |
Family
ID=83030824
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210601094.2A Pending CN114998855A (en) | 2022-05-30 | 2022-05-30 | Road edge line generation method and device, storage medium and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114998855A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117853677A (en) * | 2024-02-29 | 2024-04-09 | 腾讯科技（深圳）有限公司 | Road edge drawing method and device, electronic equipment and storage medium |
-
2022
- 2022-05-30 CN CN202210601094.2A patent/CN114998855A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117853677A (en) * | 2024-02-29 | 2024-04-09 | 腾讯科技（深圳）有限公司 | Road edge drawing method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108629231B (en) | Obstacle detection method, apparatus, device and storage medium | |
| EP3063552B1 (en) | Method and apparatus for road width estimation | |
| EP2858008B1 (en) | Target detecting method and system | |
| US20170004345A1 (en) | Survey data processing device, survey data processing method, and survey data processing program | |
| JP5980148B2 (en) | How to measure parking occupancy from digital camera images | |
| CN111932943B (en) | Dynamic target detection method and device, storage medium and roadbed monitoring equipment | |
| GB2537681A (en) | A method of detecting objects within a 3D environment | |
| CN104239867A (en) | License plate locating method and system | |
| CN103324936A (en) | Vehicle lower boundary detection method based on multi-sensor fusion | |
| CN110674705A (en) | Small-sized obstacle detection method and device based on multi-line laser radar | |
| JP2012068961A (en) | Runway estimation device and program | |
| CN111323788B (en) | Building change monitoring method and device and computer equipment | |
| CN113421432A (en) | Traffic restriction information detection method and device, electronic equipment and storage medium | |
| CN113933859A (en) | Pavement and two-side retaining wall detection method for unmanned mine card driving scene | |
| CN114998855A (en) | Road edge line generation method and device, storage medium and computer equipment | |
| CN111160132B (en) | Method and device for determining lane where obstacle is located, electronic equipment and storage medium | |
| CN112149471B (en) | Loop detection method and device based on semantic point cloud | |
| CN114241448A (en) | Method and device for obtaining heading angle of obstacle, electronic equipment and vehicle | |
| JP2007033931A (en) | Road recognition system for map generation using satellite image or the like | |
| Mitzel et al. | Real-Time Multi-Person Tracking with Time-Constrained Detection. | |
| CN111862206A (en) | Visual positioning method and device, electronic equipment and readable storage medium | |
| CN117789159A (en) | Object detection device, object detection method, and storage medium | |
| JP2012234432A (en) | Vanishing point calculation device, vanishing point calculation method and program | |
| CN116092035A (en) | Lane line detection method, lane line detection device, computer equipment and storage medium | |
| CN115330630A (en) | Point cloud data enhancement method and device for automatic driving data set construction in mining area |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |