Method for recognizing curve in advance

Abstract

The invention provides a method for recognizing a curve in advance, which is applied to an intelligent driving vehicle and comprises the following steps: step 1, acquiring ambient light information, lane line image information and front road image information around the vehicle by a camera; step 2, the image acquisition device converts the ambient light information into an image electronic signal and performs image transmission and storage on the acquired electronic signal, and an image processor performs identification processing on an image; 3, calculating the actual length of the lane line; 4, the gradient of the current driving position of the vehicle is obtained; 5, performing image analysis on the image information of the road ahead, and identifying the vehicle position and the stop line in the image; and 6, if the current weather condition is a sunny day, the length of the lane line is not greater than a preset threshold value, the front road is not a ramp, no vehicle shields the lane line in front, and no stop line exists, judging that the front road is a curve.

Description

Technical Field

[0001] The present application relates to the field of intelligent driving technology, and in particular to a method for identifying curves in advance. Background Art

[0002] ICC (Intelligent Cruise Control) can control the speed of the vehicle according to the radius of curvature of the curve so that it can pass the curve smoothly, but the detection of the curve needs to rely on high-precision maps. In the absence of high-precision map coverage, it is difficult to detect the curve information before the curve. Generally, the vehicle has to drive to the middle of the curve to detect the curve information, resulting in the curve deceleration too late, and the vehicle rushes out of the lane and cannot be centered. Therefore, in the absence of high-precision maps, it is very important to identify the curve in advance. The existing technology will be affected and limited in the face of complex road environments. For example, it may not be able to provide sufficiently accurate detection results when facing occlusion and complex road conditions. Under complex and severe weather conditions, the detection ability of the camera itself will also be reduced. Summary of the invention

[0003] In view of this, the present invention provides a method for identifying curves in advance to solve the technical problem that the prior art cannot use a camera to identify curves in advance in a complex road environment, thereby avoiding the vehicle running out of the lane and colliding due to vehicle control delay.

[0004] The present invention provides a method for identifying a curve in advance, which is applied to an intelligent driving vehicle. The method comprises: step 1, a camera collects ambient light information, lane line image information, and front road image information around the vehicle; step 2, an image acquisition device converts the ambient light information into an image electronic signal, and performs image transmission and storage on the collected electronic information, and an image processor performs recognition processing on the image to identify the current weather conditions; step 3, detecting the lane line image information by image processing technology, and calculating the actual length of the lane line according to the coordinates of the pixel points in the image; step 4, obtaining the slope of the current driving position of the vehicle according to the information obtained by the camera, combined with the information obtained by the vehicle-mounted GPS module, the high-precision map and the vehicle sensor; step 5, performing image analysis on the front road image information, and identifying the vehicle position and the stop line in the image; step 6, if the current weather condition is sunny, the length of the lane line is not greater than a preset threshold, the front road is not a ramp, there is no vehicle blocking the lane line in front, and there is no stop line, then it is determined that the front road is a curve.

[0005] Furthermore, the ambient light information includes image data of the sky and the ground.

[0006] Furthermore, the image processor performs recognition processing on the image to identify the current weather conditions, specifically: step 21, segmenting the weather image into a sky image and a ground image; step 22, extracting global features from the weather image, extracting sky features from the sky image, and extracting ground features from the ground image; step 23, fusing all features and importing them into a classifier to identify different weather conditions.

[0007] Furthermore, the weather conditions include sunny days, cloudy days, snowy days, rainy days, foggy days and dusty days.

[0008] Furthermore, the preset threshold is 50m.

[0009] Furthermore, the step 3 includes: step 31, the camera collects lane line images in real time; step 32, grayscale conversion and binarization processing are performed on the collected image information; step 33, lane line feature extraction is performed using edge detection and Hough transform and other methods; step 34, lane line detection is performed using image processing technology; step 35, the actual length of the lane line is calculated using the coordinates of the pixel points in the image.

[0010] Furthermore, the identification of the vehicle position in the road ahead image specifically includes: step 511, analyzing the road ahead image captured by the camera in real time; step 512, denoising, brightness adjustment, and color correction preprocessing of the image; step 513, using a deep learning model to analyze the image and identify the vehicle in the image; step 514, detecting and locating the target in the image, and calculating the position and size of the target.

[0011] Furthermore, the identification of the stop line in the image of the road ahead specifically includes: step 521, the vehicle-mounted camera collects the image information of the road ahead; step 522, grayscales the collected image, and sets the region of interest according to the distribution characteristics of the road traffic lines in the image; step 523, smoothes the grayscale image in the region of interest by using Gaussian filtering, and grayscale stretches the image in the region of interest to enhance the contrast of the image; step 524, calculates the grayscale gradient value of each pixel point in the region of interest in the horizontal and vertical directions, as well as the gradient direction angle, to form an effective gradient map; step 525, traverses the initial effective gradient map, removes interference points, and obtains a regional growth source map; step 526, scans the regional growth source map, Obtain an initial seed point set for regional growth, and perform regional growth to obtain a regional growth result map composed of valid points containing lane information; Step 527, perform Hough transform on the regional growth result map to obtain an initial straight line set, analyze and screen the initial straight line set to obtain an initial target straight line that meets the requirements; Step 528, determine an initial scanning area based on the initial target straight line, scan the initial scanning area, and obtain a final scanning area based on the scanning result; Step 529, scan the final scanning area for the number of valid points, calculate the percentage of the number of valid points to the total number of pixel points in the final scanning area, and when the percentage of valid points in the final scanning area is greater than or equal to the density threshold, determine that the preliminary target straight line is a valid straight line, and then identify the stop line.

[0012] Furthermore, the method also includes: step 7, after determining that the road ahead is a curve, the vehicle slows down and continues to move forward, and determines whether the road ahead is a curve in sequence until the vehicle stops moving forward.

[0013] Furthermore, the method also includes: Step 0, activating the intelligent cruise control function of the vehicle.

[0014] The present invention provides a method for identifying curves in advance, which uses a camera to collect various image information around the vehicle, and then uses an image processor, an on-board GPS module, a high-precision map, a vehicle sensor and other automotive hardware and software to calculate and identify the current weather conditions, the actual length of the lane line, the slope of the vehicle's current driving position, the position of the vehicle in front and the stop line, and finally combines the above data to determine that the road ahead is a curve. This technical solution solves the technical problem that the existing technology cannot use a camera to identify curves in advance in a complex road environment, thereby avoiding the technical problem that the vehicle runs out of the lane and collides due to vehicle control delay. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic flow chart of a method for identifying a curve in advance provided by the present invention; Figure 2It is a control process link diagram of a method for early identification of curves provided by the present invention; Figure 3 It is a schematic flow chart of another method for identifying a curve in advance provided by the present invention; Figure 4 It is a schematic diagram of the flow of the method for calculating the actual length of the lane line provided by the present invention; Figure 5 It is a schematic flow chart of the method for determining the position of a vehicle ahead provided by the present invention; Figure 6 The present invention is a flowchart of a method for identifying a stop line. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] Embodiment 1: The present invention provides a method for identifying a curve in advance, which is applied to an intelligent driving vehicle, such as Figure 1 As shown, the method steps are as follows.

[0018] Step 1: The camera collects ambient light information around the vehicle, lane line image information, and front road image information; Step 2, the image acquisition device converts the ambient light information into an image electronic signal, and transmits and stores the collected electronic information as an image, and the image processor recognizes and processes the image to identify the current weather conditions; Step 3: Detect lane line image information through image processing technology, and calculate the actual length of the lane line according to the coordinates of the pixel points in the image; Step 4, based on the information obtained by the camera, combined with the information obtained by the vehicle GPS module, the high-precision map and the vehicle sensor, the slope of the vehicle's current driving position is obtained; Step 5, performing image analysis on the image information of the road ahead to identify the vehicle position and stop line in the image; Step 6: If the current weather condition is sunny, the length of the lane line is not greater than the preset threshold, the road ahead is not a slope, there is no vehicle blocking the lane line ahead, and there is no stop line, then it is determined that the road ahead is a curve.

[0019] The present invention provides a method for identifying curves in advance, which uses a camera to collect various image information around the vehicle, and then uses an image processor, an on-board GPS module, a high-precision map, a vehicle sensor and other automotive hardware and software to calculate and identify the current weather conditions, the actual length of the lane line, the slope of the vehicle's current driving position, the position of the vehicle in front and the stop line, and finally combines the above data to determine that the road ahead is a curve. This technical solution solves the technical problem that the existing technology cannot use a camera to identify curves in advance in a complex road environment, thereby avoiding the technical problem that the vehicle runs out of the lane and collides due to vehicle control delay.

[0020] Embodiment 2: The present invention provides a method for identifying a curve in advance, which is applied to an intelligent driving vehicle, such as Figure 2 As shown, the method steps are as follows.

[0021] Step 1: The camera collects ambient light information around the vehicle, lane line image information, and front road image information; The ambient light information includes image data of the sky and the ground, and the subsequent steps will divide the image into two parts, namely, the sky and the ground, for processing.

[0022] Step 2, the image acquisition device converts the ambient light information into an image electronic signal, and transmits and stores the collected electronic information as an image, and the image processor recognizes and processes the image to identify the current weather conditions; The weather conditions include sunny days, cloudy days, snowy days, rainy days, foggy days and dusty days. The specific process of identifying the current weather conditions is as follows.

[0023] Step 21, segmenting the weather image into a sky image and a ground image; Step 22, extracting global features from the weather image, extracting sky features from the sky image, and extracting ground features from the ground image; Step 23, all features are fused and imported into the classifier to identify different weather conditions.

[0024] Since the present invention is applicable to scenes with high visibility, that is, when the weather condition is clear, the weather condition can be identified and the technical solution can be used in appropriate situations to improve safety.

[0025] Step 3: Detect lane line image information through image processing technology, and calculate the actual length of the lane line according to the coordinates of the pixel points in the image; Step 4, based on the information obtained by the camera, combined with the information obtained by the vehicle GPS module, the high-precision map and the vehicle sensor, the slope of the vehicle's current driving position is obtained; It can be seen from this step that the technical solution provided by the present invention needs to combine the vehicle-mounted GPS (Global Positioning System) module, high-precision map and vehicle sensors to obtain the provided data information to make the final judgment of the curve, rather than only analyzing the image data provided by the camera. The existing technology uses a front-view camera to detect the distance of the front lane line, monitor the front road ramp information, monitor whether the front vehicle is pressing the front lane line, identify the stop line on the road, and judge the weather conditions. However, in the face of complex road environments, the camera detection results will be affected and limited. For example, in the face of occlusion and complex road conditions, it may not be able to provide sufficiently accurate detection results, and in complex and severe weather conditions, the camera's own detection ability will also be reduced. The present invention can combine the data collected by the camera with other software and hardware on the vehicle to perform high-precision recognition and calculation, so that the vehicle can better identify the road environment, especially the curve, in a specific driving scenario.

[0026] Step 5, performing image analysis on the image information of the road ahead to identify the vehicle position and stop line in the image; Step 6: If the current weather condition is sunny, the length of the lane line is not greater than the preset threshold, the road ahead is not a slope, there is no vehicle blocking the lane line ahead, and there is no stop line, then it is determined that the road ahead is a curve.

[0027] The preset threshold described here is usually 50m, which can of course be adjusted according to actual conditions.

[0028] The present invention provides a method for identifying curves in advance, which uses a camera to collect various image information around the vehicle, and then uses an image processor, an on-board GPS module, a high-precision map, a vehicle sensor and other automotive hardware and software to calculate and identify the current weather conditions, the actual length of the lane line, the slope of the vehicle's current driving position, the position of the vehicle in front and the stop line, and finally combines the above data to determine that the road ahead is a curve. This technical solution solves the technical problem that the existing technology cannot use a camera to identify curves in advance in a complex road environment, thereby avoiding the technical problem that the vehicle runs out of the lane and collides due to vehicle control delay.

[0029] Embodiment three: The present invention provides a method for identifying a curve in advance, which is applied to an intelligent driving vehicle, such as Figure 3 As shown, the method steps are as follows.

[0030] Step 0, activate the vehicle's smart cruise control function.

[0031] The technical solution provided by the present invention is generally carried out when ICC (Intelligent Cruise Control) is activated.

[0032] Step 1: The camera collects ambient light information around the vehicle, lane line image information, and front road image information; Step 2, the image acquisition device converts the ambient light information into an image electronic signal, and transmits and stores the collected electronic information as an image, and the image processor recognizes and processes the image to identify the current weather conditions; Step 3: Detect lane line image information through image processing technology, and calculate the actual length of the lane line according to the coordinates of the pixel points in the image; like Figure 4 As shown, the specific process of calculating the actual length of the lane line is as follows.

[0033] Step 31, the camera collects lane line images in real time; Step 32, grayscale conversion and binarization processing are performed on the collected image information; Step 33, using edge detection and Hough transform methods to extract lane line features; Step 34, detecting lane lines by image processing technology; Step 35, calculating the actual length of the lane line through the coordinates of the pixel points in the image.

[0034] Step 4, based on the information obtained by the camera, combined with the information obtained by the vehicle GPS module, the high-precision map and the vehicle sensor, the slope of the vehicle's current driving position is obtained; Step 5, performing image analysis on the image information of the road ahead to identify the vehicle position and stop line in the image; like Figure 5 As shown, the specific steps of identifying the vehicle position in the front road image are as follows.

[0035] Step 511, analyzing the front road image collected by the camera in real time; Step 512, performing denoising, brightness adjustment, and color correction preprocessing on the image; Step 513, using a deep learning model to analyze the image and identify the vehicle in the image; Step 514, detect and locate the target in the image, and calculate the position and size of the target.

[0036] like Figure 6 As shown, the specific steps of identifying the stop line in the front road image are as follows.

[0037] Step 521, the vehicle-mounted camera collects image information of the road ahead; Step 522, grayscale processing is performed on the collected image, and a region of interest is set according to the distribution characteristics of the road traffic lines in the image; Step 523, using Gaussian filtering to smooth the grayscale image in the region of interest, and performing grayscale stretching on the image in the region of interest to enhance the contrast of the image; Step 524, calculating the grayscale gradient value in the horizontal and vertical directions and the gradient direction angle of each pixel point in the region of interest to form an effective gradient map; Step 525, traverse the initial effective gradient map, remove interference points, and obtain a region growing source map; Step 526, scanning the region growth source map to obtain a set of initial seed points for region growth, and performing region growth to obtain a region growth result map composed of valid points containing lane information; Step 527, performing Hough transform on the region growing result map to obtain an initial straight line set, analyzing and screening the initial straight line set to obtain an initial target straight line that meets the requirements; Step 528, determining an initial scanning area according to the initial target straight line, scanning the initial scanning area, and obtaining a final scanning area according to the scanning result; Step 529, scan the final scanning area for the number of effective points, calculate the percentage of the number of effective points to the total number of pixels in the final scanning area, and when the percentage of effective points in the final scanning area is greater than or equal to the density threshold, determine that the preliminary target straight line is a valid straight line, and then identify the stop line.

[0038] Step 6: If the current weather condition is sunny, the length of the lane line is not greater than the preset threshold, the road ahead is not a slope, there is no vehicle blocking the lane line ahead, and there is no stop line, then it is determined that the road ahead is a curve.

[0039] Step 7, after determining that the road ahead is a curve, the vehicle decelerates and continues to move forward, and determines whether the road ahead is a curve in sequence until the vehicle stops moving forward.

[0040] The present invention provides a method for identifying curves in advance, which uses a camera to collect various image information around the vehicle, and then uses an image processor, an on-board GPS module, a high-precision map, a vehicle sensor and other automotive hardware and software to calculate and identify the current weather conditions, the actual length of the lane line, the slope of the vehicle's current driving position, the position of the vehicle in front and the stop line, and finally combines the above data to determine that the road ahead is a curve. This technical solution solves the technical problem that the existing technology cannot use a camera to identify curves in advance in a complex road environment, thereby avoiding the technical problem that the vehicle runs out of the lane and collides due to vehicle control delay.

[0041] In summary, the embodiment of the present invention provides a method for identifying curves in advance, which uses a camera to collect various image information around the vehicle, and then uses the image processor, vehicle-mounted GPS module, high-precision map and vehicle sensor and other automotive hardware and software to calculate and identify the current weather conditions, the actual length of the lane line, the slope of the vehicle's current driving position, the position of the vehicle in front and the stop line, and finally combines the above data to determine that the road ahead is a curve. This technical solution solves the technical problem that the existing technology cannot use a camera to identify curves in advance in a complex road environment, thereby avoiding the technical problem that the vehicle runs out of the lane and collides due to vehicle control delays.

[0042] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for identifying a curve in advance, applied to an intelligent driving vehicle, characterized in that: The method comprises: Step 1: The camera collects ambient light information around the vehicle, lane line image information, and front road image information; Step 2, the image acquisition device converts the ambient light information into an image electronic signal, and transmits and stores the collected electronic information as an image, and the image processor recognizes and processes the image to identify the current weather conditions; Step 3: Detect lane line image information through image processing technology, and calculate the actual length of the lane line according to the coordinates of the pixel points in the image; Step 4, based on the information obtained by the camera, combined with the information obtained by the vehicle GPS module, the high-precision map and the vehicle sensor, the slope of the vehicle's current driving position is obtained; Step 5, performing image analysis on the image information of the road ahead to identify the vehicle position and stop line in the image; Step 6: If the current weather condition is sunny, the length of the lane line is not greater than the preset threshold, the road ahead is not a slope, there is no vehicle blocking the lane line ahead, and there is no stop line, then it is determined that the road ahead is a curve.

2. The method for identifying a curve in advance according to claim 1, characterized in that: The ambient light information includes image data of the sky and the ground.

3. A method for identifying a curve in advance according to claim 2, characterized in that: The image processor performs recognition processing on the image to identify the current weather conditions, specifically: Step 21, segmenting the weather image into a sky image and a ground image; Step 22, extracting global features from the weather image, extracting sky features from the sky image, and extracting ground features from the ground image; Step 23, all features are fused and imported into the classifier to identify different weather conditions.

4. A method for early identification of a curve according to claim 1 or 3, characterized in that: The weather conditions include sunny days, cloudy days, snowy days, rainy days, foggy days and dusty days.

5. The method for identifying a curve in advance according to claim 1, characterized in that: The preset threshold is 50m.

6. The method for identifying a curve in advance according to claim 1, characterized in that: The step 3 comprises: Step 31, the camera collects lane line images in real time; Step 32, grayscale conversion and binarization processing are performed on the collected image information; Step 33, using edge detection and Hough transform methods to extract lane line features; Step 34, detecting lane lines by image processing technology; Step 35, calculating the actual length of the lane line through the coordinates of the pixel points in the image.

7. The method for identifying a curve in advance according to claim 1, characterized in that: The identifying of the vehicle position in the front road image specifically includes: Step 511, analyzing the front road image collected by the camera in real time; Step 512, performing denoising, brightness adjustment, and color correction preprocessing on the image; Step 513, using a deep learning model to analyze the image and identify the vehicle in the image; Step 514, detect and locate the target in the image, and calculate the position and size of the target.

8. The method for identifying a curve in advance according to claim 1, characterized in that: The identifying a stop line in the front road image specifically includes: Step 521, the vehicle-mounted camera collects image information of the road ahead; Step 522, grayscale processing is performed on the collected image, and a region of interest is set according to the distribution characteristics of the road traffic lines in the image; Step 523, using Gaussian filtering to smooth the grayscale image in the region of interest, and performing grayscale stretching on the image in the region of interest to enhance the contrast of the image; Step 524, calculating the grayscale gradient value in the horizontal and vertical directions and the gradient direction angle of each pixel point in the region of interest to form an effective gradient map; Step 525, traverse the initial effective gradient map, remove interference points, and obtain a region growing source map; Step 526, scanning the region growth source map to obtain a set of initial seed points for region growth, and performing region growth to obtain a region growth result map composed of valid points containing lane information; Step 527, performing Hough transform on the region growing result map to obtain an initial straight line set, analyzing and screening the initial straight line set to obtain an initial target straight line that meets the requirements; Step 528, determining an initial scanning area according to the initial target straight line, scanning the initial scanning area, and obtaining a final scanning area according to the scanning result; Step 529, scan the final scanning area for the number of effective points, calculate the percentage of the number of effective points to the total number of pixels in the final scanning area, and when the percentage of effective points in the final scanning area is greater than or equal to the density threshold, determine that the preliminary target straight line is a valid straight line, and then identify the stop line.

9. The method for identifying a curve in advance according to claim 1, characterized in that: The method further comprises: Step 7, after determining that the road ahead is a curve, the vehicle decelerates and continues to move forward, and determines whether the road ahead is a curve in sequence until the vehicle stops moving forward.

10. The method for identifying a curve in advance according to claim 1, characterized in that: The method further comprises: Step 0, activate the vehicle's smart cruise control function.

Images