JPH0523298U - Roadway judging device - Google Patents

Abstract

(57) [Abstract] [Purpose] To enable high-precision judgment of a motorway at all times when needed and without being influenced by regional characteristics. [Structure] An image recognition processing unit 4 extracts a feature of a lane marking on a traveling motorway, and a judging unit 6 determines whether the feature-carrying motorway is a highway or another Determine if it is a general motorway.

Description

[Detailed description of the device]

[0001]

[Industrial applications]

 The present invention relates to a motorway determination device that determines whether a traveling motorway is a highway (including an equivalent motorway) or a general motorway other than that.

[0002]

[Prior Art]

 Since the road design is different between the highway and the general motorway, and the legal speeds are also different, it is necessary to judge whether the traveling motorway is the highway or the general motorway. To determine such a motorway, there are those that use the map information of the navigation system or that receive information from ground equipment.

[0003]

[Problems to be solved by the device]

 In the navigation system method, it is necessary to determine the position of the vehicle with high accuracy, but the accuracy of the determination is not always high because the current satellite is used, which is not always stable in operation at this stage. There is a time zone in which it is not possible to determine the vehicle position. In addition, in the method using the ground equipment, there are areas where the ground equipment does not exist, so that the determination remains regional.

Therefore, an object of the present invention is to make it possible to judge a motorway with high accuracy when necessary and without being influenced by regional characteristics.

[0005]

[Means for Solving the Problems]

 In order to achieve such an object, the present invention has an image recognition processing means and a determination means, and is mounted on a vehicle, and the image recognition processing means is a vehicle that runs on the vehicle. The lane markings of the existing lane markings are image-recognized, and the features of the lane markings that have been image-recognized are extracted, and the judging means determines that the traveling lane is a high-speed vehicle based on the extracted features of the lane markings. It is characterized in that it determines whether it is a road (including a motorway according to this) or an ordinary motorway. Also, this lane marking is a marking that consists of repeated lines and blanks, and the feature to be extracted from the lane marking is the length of the line, the length of the blank, or the repeating pitch thereof. It may be at least one. Further, the lane marking is a marking consisting of repeated lines and blanks, and the determining means determines the length of the line, the length of the blank, and the repeating pitch, respectively, as features to extract the lane markings. The antecedent part variable is defined as the antecedent part variable, the motorway judgment is the consequent part variable, and the judgment is made by fuzzy inference according to the fuzzy rule composed of both variables and the membership function defined corresponding to each variable. May be done.

[0006]

[Action]

 The image recognition processing means image-recognizes the lane markings on the motorway on which the vehicle is traveling, and extracts the features of the lane markings that have been image-recognized. Then, the determining means determines whether the traveling highway is a highway (including a highway corresponding to this) or a general highway based on the characteristics of the extracted lane markings.

In the case where this lane marking is a marking consisting of repeated lines and blanks, the feature to be extracted from the lane marking is that the length of the line, the length of the blank, or the repetition pitch thereof. It can be at least one. Further, by performing the determination by the determination means by fuzzy inference, accurate determination can be performed even if the lane marking line or blank is not clear in image recognition.

[0008]

【Example】

 Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a circuit block diagram of the device of the present embodiment. As shown in FIG. 1, the device 2 of the present embodiment has an image recognition processing unit 4 and a determination unit 6.

The image recognition processing unit 4 is for recognizing an image of a lane marking on a driving road, and has a photographing unit 4a, a recognition unit 4b, and a calculation unit 4c. The road is captured, and the captured image is recognized by the recognition unit 4b with a change in brightness as shown in FIG. The hatching in FIG. 2 is the white line of the lane marking, and is the part where the brightness is higher than the part other than the hatching. FIG. 2 shows a recognition image 8 of a road ahead of a moving vehicle which is photographed by the photographing unit 4a and recognized by the recognition unit 4b. On the right side of the recognition image 8 is a white line 10 indicating continuous lane markings. Also, white lines 12 indicating broken lane markings are shown on the left side of the recognition image 8, respectively, but the white lines 10 and 12 have high brightness, and the white line 12 between the white lines 12 on the other side is a blank space. The brightness is low, including. Then, the recognition unit 4b sets the window 16 at the lower center of the recognized image 8 (the road surface in front of the automobile), obtains the average luminance in the window 16 at the lower center as the reference average luminance K, and First, the comparative average luminance K'at each window position is obtained by shifting 16 to the right end of the recognized image 8 in the arrow direction, and the comparative average luminance K'at the right white line 10 exceeds the reference average luminance K. By doing so, it is recognized that there is a white line 10 on the right side of the lane marking, and then the comparative average brightness K ′ at each window position is obtained by shifting the window 16 to the left end of the recognized image 8 in the direction of the arrow. If the comparative average luminance K ′ at 12 exceeds the reference average luminance K, the white line 12 on the left side of the lane marking and the blank 14 between the white lines 12 on the left side However if the comparison average luminance K 'is the reference average brightness K or less value, if space between the left side white line 12 of the lane markings 14 are present, respectively, recognized. In this way, the left and right lane markings of the recognition image 8 are recognized by shifting the window 16 to the left and right ends of the recognition image 8 and shifting it upward. After recognizing the lane markings in this way, in order to make data for determining whether this lane marking is that of a high-speed vehicle or that of a general vehicle, the calculation unit 4c uses the lane marking on the left side as the data. , The length L1 of the white line 12, the length L2 of the blank 14, and the repeating pitch P of the white line 12 and the blank 14, respectively. For this, a well-known coordinate conversion method in image recognition is used. It is possible to obtain it by adopting it. For example, the coordinate points on the recognition image 8 at the points a, b, and c are obtained, and the distance between the points a and b at the coordinate points on the recognition image 8 and the distance between the points b and c. Converting the distances into actual distances on the road, the distance between points a and b is the blank length L2 in the lane marking, and the distance between point b and c is the white line length L1 in the lane marking. Further, the repeating pitch P thereof is also obtained at the same time. Then, the calculating unit 4c outputs the white line length L1, the blank length L2, and the repeating pitch P thus obtained to the judging unit 6 as the feature extraction data of the lane marking for judging the road on which the vehicle is running. ..

In addition, in FIG. 2, the brightness change of the recognition image 8 is used. However, in the recognition unit 4b in the image recognition processing unit 4, the edges of the white lines 10 and 12 on the right side and the left side of the lane markings, respectively. When the differential processing is performed at that point, the lane marking captured by the image capturing unit 4a becomes a differential processing recognition image as shown in FIG. In FIG. 3, the left edges 10 'and 12' of the white lines 10 and 12 respectively on the right and left sides are solid lines, and the right edges 10 '' and 12 '' are dashed lines, respectively. The solid lines of 10 'and 12' show the rising of the differentiation, and the broken lines of the right edges 10 '' and 12 '' show the falling of the differentiation. Then, the absolute value of the magnitude of the rise and fall of the differential value is compared with the reference value, and if the absolute value exceeds the reference value, it is the white line of the lane marking, and if it does not exceed it, there is it. Is recognized as a blank lane marking. In this way, the white line length L1, the blank length L2, and the repeating pitch P are obtained from the recognized lane markings by the same calculation unit 4c as described above, and the determination unit 6 determines the lane markings of the traveling lane. Output as feature extraction data.

The determination unit 6 uses the lane marking feature extraction data from the image recognition processing unit 4 as the white line length, the blank length, and the white line length in the lane marking standardized for each of the high-speed automatic road and the general highway. Compare with the features of normalized lane markings for repeat pitch. This standardized lane marking is based on the regulations based on the Ordinance of the Prime Minister's Office / Ministry of Construction “Order for Road Signs / Division Lines and Road Marks”. When the repeating pitch is P = 20 m (8 m + 12 m), in the case of a general motorway, the white line length L1 = 6 m, the blank length L2 = 9 m, and the repeating pitch P = 15 m (6 m + 9 m).

Then, the determination unit 6 determines from the comparison whether the traveling motorway is a highway (including a motorway conforming to it) or another general motorway. It is supposed to do.

The determination unit 6 makes the determination by fuzzy inference. That is, the determination unit 6 is provided with the feature extraction data of the lane markings regarding the white line length L1, the blank length L2, and the repeating pitch P from the image recognition processing unit 4. The determination unit 6 sets each of these data L1, L2, P as an antecedent part variable, sets the motorway determination as a consequent part variable Y, and corresponds to each fuzzy rule and each variable. According to the membership function, the judgment is made by fuzzy inference by the MIN-MAX AND method. A typical example of this fuzzy rule is shown below.

If L1 = “unknown” and L2 = “general” and P = “general” then Y = “general motorway” if L1 = “unknown” and L2 = “highway” andP = “general” then Y = “general motorway” If L1 = “general” and L2 = “general” and P = “general” then Y = “general motorway” if L1 = “general” and L2 = “high speed” andP = “high speed” then Y = “high speed motorway” here Where if is the conditional part (preceding part), then is the conclusion part (consequent part), L1 is the antecedent variable of the white line length, L2 is the antecedent part variable of the blank length, and P is the white line. Blank repeat pitch antecedent part variable, Y is consequent part variable of judgment. In addition, "Unknown" is unclear whether it is a general highway or a highway, "General" is that the highway is a general highway, "Highway" is that the highway is a highway, "Highway" is a fuzzy label indicating that the determination is a general highway, and "Highway" is a determination that is a highway. The membership function of the antecedent variable for each of the white line length, the blank length, and the repeating pitch and the membership function of the antecedent variable for the determination are shown in FIG. 4, respectively.

The determination unit 6 uses fuzzy inference according to the so-called MIN-MAX logical product method. First, the minimum value of the membership value of each antecedent variable is calculated for each fuzzy rule ,. It is selected as the antecedent fitness using the membership function of 5 (MIN operation). Then, the membership function shown in FIG. 4 of the consequent part variable Y for each antecedent part suitability for each fuzzy rule is cut. By this cutting, the membership value of the consequent part variable for each fuzzy rule is obtained, and at the same time, the one with the largest membership value is selected (MAX operation). Since such fuzzy inference is well known, its details are omitted.

A control reference selection unit is connected to the latter stage of the determination unit 6, a vehicle control unit is further connected to the latter stage thereof, and the control reference selection unit determines whether the vehicle is a general highway or a highway based on the determination output. Alternatively, the vehicle control unit may control the vehicle according to the control standard selected in accordance with the above. Even if the vehicle is traveling at the same speed, the control reference selection section is designed to match the curve curvature and the frequency of acceleration between a general highway and a highway, so that the vehicle should be designed accordingly. Therefore, the control reference is selected in response to the judgment output, and the vehicle control unit is driven by this.

Further, a maximum speed determination unit and a vehicle speed control unit are connected downstream of the determination unit 6, and the maximum speed determination unit responds to the roadway determination output from the determination unit 6 to determine whether the vehicle type is the own vehicle type. The maximum speed on the motorway may be estimated, and the vehicle speed control unit may control the speed of the vehicle according to the estimation.

[0019]

[Effect of the device]

 As is apparent from the above description, according to the present invention, the image recognition processing means extracts the characteristics of the lane markings on the traveling road, and the characteristics of the lane markings are given to the judging means. Because of the characteristics of the judging means, it is judged whether the motorway being driven is a highway or another general motorway. It became possible to judge the motorway without being affected by regional characteristics.

Further, if the line length of the lane marking, the blank length, or their repeated pitch is used as the feature extraction of the lane marking, the standards are different between the highway and the general highway. The determination by the determination means becomes easy. Furthermore, by performing the determination by fuzzy inference, even if the lane markings on the traveling highway are unclear, the highway can be accurately determined.

[Brief description of drawings]

FIG. 1 is a circuit block diagram of an apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing an image in a case where a recognition image in front of a vehicle is recognized by brightness by an image recognition processing unit of the same apparatus.

FIG. 3 is a diagram showing an image in the case of recognition by differentiating the recognition image of the same apparatus.

FIG. 4 is a diagram showing a membership function necessary for fuzzy inference when the determination unit performs fuzzy inference.

[Explanation of symbols]

 4 Image recognition processing unit 6 Judgment unit 8 Recognition image 10, 12 White line 14 Blank

Claims (3)

[Scope of utility model registration request] 1. An image recognition processing means and a determination means, which are mounted on a vehicle, wherein the image recognition processing means recognizes an image of a lane marking of an automobile road on which the vehicle is traveling. The feature of the lane marking that has been image-recognized is extracted, and the determination means is a highway (including a corresponding highway) where the traveling highway is based on the extracted features of the lane marking. A roadway determination device for determining whether the vehicle is a general roadway or a general roadway. 2. The lane marking is a marking consisting of repeated lines and blanks, and the feature to be extracted from the lane marking is one of the length of the line, the length of the blank, or their repeating pitch. The roadway determination device according to claim 1, wherein the roadway determination device is at least one of the above. 3. The lane marking is a marking made up of repeated lines and blanks, and the determining means determines the length of the line, the length of the blank, and
The repeated pitch is used as the antecedent variable that is the feature to be extracted from the lane markings, and the roadway judgment is used as the antecedent variable. The vehicle road determination device according to claim 1, wherein the determination is performed by fuzzy inference according to a membership function determined by the above.