JP6476927B2 - Detection apparatus, detection method and detection program - Google Patents

Description

  The present invention relates to a detection device and the like.

  There is a prior art 1 that detects a vehicle division line from an image captured by a monocular camera installed in a vehicle. In the following description, vehicle division lines are simply referred to as white lines. The prior art 1 detects a white line using the fact that the brightness of the white line is higher than that of the road surface. For example, in the prior art 1, an image is scanned in a direction orthogonal to the traveling direction of the vehicle, and a change in luminance becomes low → high → low, and a line element following the traveling direction of the vehicle is extracted as an edge. Extract as a white line candidate. For example, an LDW (lane departure warning) apparatus or the like mounted on a vehicle outputs an alarm or the like when the vehicle is about to deviate from the white line, using the extraction result of the white line candidate.

  However, when the white line candidate is extracted by the prior art 1, there is a case where the road surface paint is erroneously extracted as the white line candidate. For example, it is assumed that the road surface paint indicates a restriction mark painted on the road surface and an indication mark excluding a vehicle division line. FIG. 19 is a diagram for explaining the problem of the prior art 1. As shown in FIG. 19, for example, when a line element whose brightness change is low → high → low and continues in the traveling direction of the vehicle is extracted as an edge, the white line candidate 10a is extracted. 10a is a part of the road surface paint and is not a white line candidate. For this reason, when the prior art 1 is used, the road surface paint may be erroneously detected as a white line candidate, and the LDW device can not execute correct deviation determination.

  The prior art 2 exists to solve the problems of the prior art 1. FIG. 20 is a diagram for explaining the prior art 2. The prior art 2 stores road surface paint information as dictionary data in the database 11 in advance, and compares the extracted white line candidates with the dictionary data to exclude white line candidates that become part of the road surface paint from the white line candidates. Do. For example, in FIG. 20, since the white line candidate 10a matches the road surface pattern 11a included in the dictionary data of the database 11, the prior art 2 excludes the white line candidate 10a from the white line candidates.

JP 2012-150595 A JP 2003-85562 A JP, 2009-237901, A

  However, in the above-described prior art, there is a problem that only white lines can not be extracted without misjudgement from various road markings painted on a road surface.

  For example, in the prior art 2 described above, since it is determined whether the white line candidate is a white line candidate using the dictionary data stored in the database 11, the road surface paint not existing in the dictionary data can not be coped with. Further, in the prior art 2, since the matching process is performed according to the number of road surface paints registered in the dictionary data, there is a problem that the amount of calculation increases as the number of road surface paints increases.

  In one aspect, it is an object of the present invention to provide a detection device, a detection method, and a detection program capable of extracting only white lines from various road markings painted on a road surface without misjudgment.

  In a first solution, the detection device includes an extraction unit and a detection unit. The extraction unit extracts a first line element continuous in the traveling direction of the vehicle as a white line candidate from the image information. In the image information, when there is a second line element having a direction different from that of the first line element and intersecting the first line element, the detection unit has the luminance feature of the first line element and the second line element. Whether or not they are the same is determined, and when the luminance features of the first line element and the second line element are the same, they are excluded from the white line candidates and a white line is detected.

  According to one embodiment of the present invention, only white lines can be extracted without misjudgement from various road markings painted on the road surface.

FIG. 1 is a functional block diagram showing the configuration of the LDW device according to the present embodiment. FIG. 2 is a diagram for explaining the common feature confirmation process. FIG. 3 is a diagram showing an example of multiple routes. FIG. 4 is a figure for demonstrating the problem of common feature recognition processing. FIG. 5 is a diagram (1) for explaining the validation process. FIG. 6 is a diagram (2) for explaining the validation process. FIG. 7 is a diagram showing an example of the traveling direction marking. FIG. 8 is a diagram (1) for explaining the establishment confirmation process. FIG. 9 is a diagram (2) for explaining the feasibility confirmation process. FIG. 10 is a diagram (3) for explaining the feasibility confirmation process. FIG. 11 is a diagram (4) for explaining the feasibility confirmation process. FIG. 12 is a diagram (5) for explaining the feasibility confirmation process. FIG. 13 is a flowchart of the white line candidate detection process. FIG. 14 is a flowchart of the common feature confirmation process. FIG. 15 is a flowchart showing the procedure of the validation process. FIG. 16 is a flowchart showing the procedure of the establishment confirmation process. FIG. 17 is a flowchart showing the processing procedure of the LDW device. FIG. 18 is a diagram illustrating an example of a computer that executes a detection program. FIG. 19 is a diagram for explaining the problem of the prior art 1. FIG. 20 is a diagram for explaining the prior art 2.

  Hereinafter, embodiments of a detection device, a detection method, and a detection program disclosed in the present application will be described in detail based on the drawings. The present invention is not limited by this embodiment.

  An example of the configuration of the LDW device according to the present embodiment will be described. FIG. 1 is a functional block diagram showing the configuration of the LDW device according to the present embodiment. As shown in FIG. 1, the LDW device 100 is connected to the imaging device 50. The LDW device 100 includes an extraction unit 110, a detection unit 120, a deviation determination unit 130, and an alarm unit 140. For example, the LDW device 100 is mounted on a vehicle (not shown). The LDW device 100 is an example of a detection device.

  The imaging device 50 is a camera mounted on a vehicle so as to capture an image of a road in a front area or a rear area of the vehicle. The imaging device 50 outputs the information of the captured image to the LDW device 100. For example, the imaging device 50 captures an image every predetermined time, and outputs information of the captured image to the LDW device 100. In the following description, image information is referred to as image information.

  The extraction unit 110 is a processing unit that extracts, based on the image information acquired from the imaging device 50, the first line element continuous in the traveling direction of the vehicle. For example, the extraction unit 110 extracts edge information from image information using a Laplacian of Gaussian (LOG) filter. The extraction unit 110 performs labeling on edge information, removes noise components in consideration of the direction and shape of line elements, and executes connection processing to extract line elements that continue in the traveling direction of the vehicle. Do.

  When the extracted line element satisfies the following condition 1 and condition 2, the extraction unit 110 sets the extracted line element as a first line element. Condition 1 is a condition that the relationship between the brightness of the area on the left side of the line element, the brightness of the line element, and the brightness of the area on the right side of the line element is less than a predetermined brightness, greater than a predetermined brightness and less than a predetermined brightness. . Condition 2 is the condition that the width of the area surrounded by each line element is the same as the width of the vehicle division line.

  The extraction unit 110 outputs the information of the first line element and the information of the other line elements to the detection unit 120. For example, the line element information includes the start point and the end point of the line element.

  The detection unit 120 is a processing unit that detects a white line candidate from the first line element. The detection unit 120 performs “common feature confirmation processing”, “validity confirmation processing”, and “certainty confirmation processing” to determine whether the first line element is recognized as a white line candidate or not. The detection unit 120 outputs the information on the first line element identified as the white line candidate to the deviation determination unit 130.

  The “common feature confirmation process” performed by the detection unit 120 will be described. FIG. 2 is a diagram for explaining the common feature confirmation process. The common feature confirmation process focuses on the following geometrical facts that the road surface paint has in common. That is, the road surface paint including the line element erroneously detected as the white line candidate is configured by a first line element continuing in the traveling direction and a second line element continuing in the direction different from the traveling direction. In addition, as far as one road surface paint is concerned, the first line element and the second line element constituting the road surface paint have the same luminance feature. For example, the brightness feature of the first line element is the brightness value on the first line element, and the brightness feature of the second line element is the brightness value on the second line element.

For example, when the relationship between the luminance value V1 on the first line element and the luminance value V2 on the second line element satisfies the expression (1), the detection unit 120 detects the luminance feature of the first line element, It is determined that the luminance feature of the second line element is the same. Vth of equation (1) is a predetermined threshold.
| V1-V2 | <Vth (1)

  For example, the detection unit 120 sets the average luminance of the first line element from which the influence of chattering near the edge and the noise component have been eliminated as the luminance value V1. The detection unit 120 sets the average luminance of the second line element from which the influence of chattering near the edge and the noise component have been eliminated as the luminance value V2.

  Arrow 5 in FIG. 2 is an arrow indicating the traveling direction of the vehicle. The line elements 12a, 12b and 12c are first line elements continuing in the traveling direction. The line elements 13a, 13b, 13c, 13d, 13e and 13f are second line elements that continue in a direction different from the traveling direction. Here, it is assumed that the first line elements 12a and 12b intersect the second line elements 13a and 13b, and the luminance features of the first line elements 12a and 12b and the luminance features of the second line elements 13a and 13b are the same. The first line elements 12a and 12b can be said to be part of the road surface paint.

  In FIG. 2, assuming that the first line element 12 c intersects with the second line elements 13 c to 13 f and the luminance feature of the first line element 12 c and the luminance feature of the second line elements 13 c to 13 f are the same, The element 12c can be said to be part of the road surface paint.

The detection unit 120 estimates the first line element as a road surface paint when the first line element satisfies the following conditions A1 and A2. If the first line element does not satisfy the condition A1 or the condition A2, the detection unit 120 recognizes the first line element as a white line candidate.
Condition A1: There is a second line element having a direction different from that of the first line element and intersecting the first line element.
Condition A2: The luminance feature of the first line element and the luminance feature of the second line element are the same.

  Here, in the actual traveling environment, if the white line candidate is identified by performing only the above-mentioned common feature confirmation processing, it may not be possible to extract multiple routes such as a decelerating road marking road. In the following description, the decelerating road marking road will be appropriately described as a multiple route. It is desirable to extract the first line element corresponding to the multiple route as a white line candidate. FIG. 3 is a diagram showing an example of multiple routes. For example, the multiplex route has a broken line 14a and a short broken line 14b.

  FIG. 4 is a figure for demonstrating the problem of common feature recognition processing. When the resolution of the imaging device 50 is low, as shown in FIG. 4, there may be a case where the fusion region 15 is formed by fusing the broken line 14 a of multiple routes and the short broken line 14 b on the image. Since the first line element included in the fused area 15 satisfies the above conditions A1 and A2, the first line element may not be recognized as a white line candidate by the common feature confirmation process described above.

  In order to solve the problem of the common feature confirmation process described with reference to FIG. 4, the detection unit 120 executes the following validity confirmation process for the first line element estimated to be part of the road surface paint.

  The “validity confirmation process” performed by the detection unit 120 will be described. 5 and 6 are diagrams for explaining the validation process. The validation process focuses on the geometrical difference between the road surface paint including the first line element excluded from the white line candidate and the multiple route including the first line element identified as the white line candidate. As shown in FIG. 5, the road surface paint including the first line element excluded from the white line candidate is widely present in the direction orthogonal to the traveling direction. For example, the width of the road surface paint 16a is about 1.2 m, and the width of the road surface paint 16b is about 1.5 m. On the other hand, as shown in FIG. 6, the width of the multiplex line 17 is about 75 cm at maximum.

  The detection unit 120 identifies a region including the first line element estimated to be road surface paint and the second line element intersecting the first line element by common feature recognition processing, and the width of the specified area should be excluded. Evaluate whether it is appropriate as a road surface paint. Specifically, when the area width W is less than the predetermined threshold value Wth, the detection unit 120 estimates that the first line element included in the area is a part of the multiplex route. For example, although the predetermined threshold value Wth is set to 75 cm, it is not limited to this. On the other hand, when the area width W is equal to or more than the predetermined threshold value Wth, the detection unit 120 determines that the first line element included in the area is a part of the road surface paint. As described above, the validity check process is performed on the first line element determined to be part of the road surface paint by the common feature recognition process, thereby preventing the detection of the first line element corresponding to multiple routes. can do.

  Here, when the common feature confirmation processing and the validity confirmation processing are performed and it is determined that the first line element is a part of the multiple route, the first line element included in the traveling direction marking is not the multiple route. , And may be misjudged as part of multiple routes. The traveling direction marking is a road surface paint, and it is not preferable to identify the first line element which is a part of the traveling direction marking as a white line candidate.

  FIG. 7 is a diagram showing an example of the traveling direction marking. For example, the width of the traveling direction mark 18 is as narrow as about 40 cm to 60 cm. Therefore, for example, when the detection unit 120 performs the common feature confirmation processing and the validity confirmation processing and estimates that the first line element is a part of the multiple route is a part of the traveling direction marking. There is. In order to solve this problem, the detection unit 120 executes the establishment confirmation process on the first line element estimated to be a part of the multiplex route by the above process.

  The “success check process” performed by the detection unit 120 will be described. 8 to 12 are diagrams for explaining the establishment confirmation process. As shown in FIG. 8, when the first wire element 20a is a part of a multiplex route, the distance between the first wire element 20a and the vehicle 5a is global even if the vehicle 5a moves in the traveling direction. It becomes constant (distance D) when observed. For example, when the distance between the first element and the vehicle is not strictly constant when observed in time series, it fluctuates with a minute amplitude (about several millimeters) due to the influence of the play of the steering wheel, road surface noise, etc. There is. For example, the detection unit 120 performs statistical processing such as smoothing on the result of continuous plural time observation, and when viewed from a broad perspective, if the first line element 20a is a part of multiple routes, the first line The distance between the element 20a and the vehicle 5a seems to be constant. Using this, the detection unit 120 divides the road surface paint into white line candidates. In FIG. 9, the relationship between the distance between the vehicle 5a and the first wire element 20a and the time is shown by a graph. In the graph of FIG. 9, the vertical axis indicates distance, and the horizontal axis indicates time. As shown in FIG. 9, when the distance is generally observed with respect to time and becomes constant, the first line element 20a is determined to be a part of the multiplex route.

  On the other hand, as shown in FIG. 10, when the first wire element 20b is a part of the traveling direction marking, when the vehicle 5a moves in the traveling direction, the distance between the first wire element and the vehicle 5a is It does not become constant by observing it on a global basis. In the example shown in FIG. 10, the distance between the first line element and the vehicle 5a is converted as distance D → distance D '→ distance D. The traveling direction marking does not last long in the traveling direction, but is isolated at the center of the lane on the road where the lane markings exist. Therefore, when the distance between the first wire element and the vehicle 5a is observed in time series, the distance changes when the first wire element included in the traveling direction marking is extracted, and the distance between the first wire element and the vehicle However, it is not constant because it is observed globally. In FIG. 11, the relationship between the distance between the vehicle 5a and the first line element and the time is shown by a graph. In FIGS. 10 and 11, in the time zone in which the first line element 20a is detected, the distance between the first line element 20a and the vehicle 5a is the distance D, and in the time zone in which the first line element 20b is detected. The distance between the first line element 20b and the vehicle 5a is the distance D '. Therefore, the first line element 20b is identified as the first line element included in the traveling direction marking.

  That is, the detection unit 120 performs the common feature confirmation process and the validity confirmation process, and measures the distance between the first line element and the vehicle for the first line element estimated to be part of the multiple route, It is determined whether the distance is constant for a predetermined time. If the distance between the first element and the vehicle becomes constant for a predetermined time, the detection unit 120 recognizes the first element as a white line candidate. On the other hand, when the distance between the first line element and the vehicle is not constant for a predetermined time, the detection unit 120 excludes the first line element from the white line candidates.

  The detection unit 120 may measure the distance between the first wire element and the vehicle in any manner. For example, the detection unit 120 uses the conversion table for converting the distance between the image information center line and the first line element into the actual distance between the vehicle and the first line element. Measure the distance. In addition, the detection unit 120 may store the distance between the first element and the vehicle measured from the past image information in a database (not shown). The detection unit 120 compares the distance between the first line element measured from the current image information and the vehicle with the past distance stored in the database, and determines that the distance between the first line element and the vehicle is a predetermined time. In the meantime, it is observed globally to determine whether or not it becomes constant.

  By the way, in a scene where there are multiple lines, the detection unit 120 pays attention to the presence of vehicle division lines on both the left and right sides of the road, and whether the road width of the first line element and other white line candidates is equal to the road width Whether or not the first line element is a white line candidate may be determined based on For example, as shown in FIG. 12, it is assumed that the first line element 20c exists on the left side of the road 40 and the white line candidate 30 exists on the right side. Then, when the distance between the first line element 20c and the white line candidate 30 is equal to the preset road width, the first line element 20c is recognized as a white line candidate.

  The detection unit 120 executes the common feature confirmation processing, the validity confirmation processing, and the validity confirmation processing described above, and outputs the information on the first line element identified as the white line candidate to the deviation determination unit 130.

  The departure determination unit 130 is a processing unit that determines whether the vehicle has deviated from the vehicle division line based on the position of the vehicle and the position of the white line candidate. For example, when the image information vehicle division line crosses the center line of the image information, the departure determination unit 130 determines that the vehicle has deviated from the vehicle division line. For example, the departure determining unit 130 determines that the vehicle has deviated from the vehicle parting line when the vehicle parting line existing on the left side of the vehicle parting line moves to the right side of the center line. Alternatively, the departure determining unit 130 determines that the vehicle has deviated from the vehicle parting line when the vehicle parting line existing on the right side of the vehicle parting line moves to the left side of the center line.

  Next, an example of the white line candidate detection process performed by the LDW device 100 according to the present embodiment will be described. FIG. 13 is a flowchart of the white line candidate detection process. For example, the process shown in FIG. 13 is repeatedly executed each time the LDW apparatus acquires image information.

  As illustrated in FIG. 13, the extraction unit 110 of the LDW device 100 acquires image information from the imaging device 50 (step S101), and extracts a first line element (step S102).

  The detection unit 120 of the LDW device 100 determines whether the first line element is present (step S103). If the first line element does not exist (step S103, No), the detection unit 120 ends the white line candidate detection process.

  On the other hand, when the first line element exists (Yes at Step S103), the detection unit 120 executes the common feature confirmation process (Step S104). If the detection unit 120 estimates that the first line element is the road surface paint as a result of executing the common feature confirmation process (Yes at step S105), the process proceeds to step S107. On the other hand, if the detection unit 120 does not estimate that it is the first line element paint as a result of executing the common feature confirmation process (step S105, No), the first line element is identified as a white line candidate (step S105) S106), the white line candidate detection process is ended.

  The detection unit 120 executes the validity confirmation process (step S107). If the detection unit 120 determines that the first line element is the road surface paint as a result of executing the validity confirmation process (step S108, Yes), the first line element is excluded from the white line candidates (step S109). , White line candidate detection processing ends. On the other hand, when the detection unit 120 estimates that the first line element is not a road surface paint as a result of executing the validity confirmation process (No in step S108), the process proceeds to step S110.

  The detection unit 120 executes the establishment confirmation process (step S110). If the detection unit 120 determines that the first line element is the road surface paint as a result of executing the establishment confirmation process (Yes at step S111), the first line element is excluded from the white line candidates (step S112). , White line candidate detection processing ends. On the other hand, when the detection unit 120 does not recognize the first line element as the road surface paint as a result of executing the establishment confirmation process (No at Step S111), the detection unit 120 recognizes the first line element as a white line candidate (Step S113). , White line candidate detection processing ends.

  Next, the procedure of the common feature confirmation process shown in step S104 of FIG. 13 will be described. FIG. 14 is a flowchart of the common feature confirmation process. As illustrated in FIG. 14, the detection unit 120 determines whether there is a second line element that intersects the first line element (step S201). If the second line element intersecting the first line element does not exist (step S202, No), the detection unit 120 determines that the first line element is a white line candidate (step S203), and performs common feature confirmation processing. finish.

  On the other hand, when there is a second line element that intersects the first line element (Yes at step S202), the detection unit 120 calculates the luminance feature of the first line element (step S204). The detection unit 120 calculates the luminance feature of the second line element (step S205).

  The detection unit 120 determines whether or not the luminance features of the first line element and the second line element are the same (step S206). If the luminance features of the first line element and the second line element are not the same (step S206, No), the detection unit 120 proceeds to step S203.

  On the other hand, when the luminance features of the first line element and the second line element are the same (Yes at step S206), the detection unit 120 determines that the first line element is road surface paint (step S207). , End the common feature confirmation process.

  Next, the process procedure of the validity check process shown in step S107 of FIG. 13 will be described. FIG. 15 is a flowchart showing the procedure of the validation process. As shown in FIG. 15, the detection unit 120 determines a road surface paint area including the first line element and the second line element (step S301). The detection unit 120 determines whether the area width W of the road surface paint area is equal to or greater than a predetermined threshold value Wth (step S302).

  If the area width W of the road surface paint area is equal to or greater than the predetermined threshold value Wth (Yes at step S302), the detection unit 120 determines that the first line element is the road surface paint (step S303). End the confirmation process.

  On the other hand, when the area width W of the road surface paint area is smaller than the predetermined threshold value Wth (No in step S302), the detection unit 120 ends the validity confirmation process.

  Next, the processing procedure of the validity confirmation process shown in step S110 of FIG. 13 will be described. FIG. 16 is a flowchart showing the procedure of the establishment confirmation process. As shown in FIG. 16, the detection unit 120 measures the position of the first line element (step S401).

  The detection unit 120 evaluates the position of the first line element (step S402), and determines whether the first line element is suitable as a lane boundary line (step S403). In steps S402 and S403, the detection unit 120 measures the distance between the first wire element and the vehicle, and the first wire element is in the lane when the distance is globally observed for a predetermined time and becomes constant. Determined to be suitable as a boundary line. On the other hand, the detection unit 120 measures the distance between the first line element and the vehicle, and if the distance is not globally observed for a predetermined time and does not become constant, then the first line element serves as a lane boundary. It determines that it is not suitable.

  If the detection unit 120 determines that the first line element is not suitable as a lane boundary line (step S403, No), the detection unit 120 determines that the first line element is a road surface paint (step S404), and the feasibility check processing Finish.

  On the other hand, when the detection unit 120 determines that the first line element is suitable as a vehicle boundary (Yes at step S403), the detection unit 120 determines whether a white line candidate exists on the opposite side of the corresponding first line element. (Step S405). If a white line candidate does not exist on the opposite side of the corresponding first line element (No at Step S405), the detection unit 120 proceeds to Step S404.

  On the other hand, when a white line candidate exists on the opposite side of the corresponding first line element (Yes in step S405), the detection unit 120 corresponds to the road width between the first line element and the white line candidate on the opposite side. It is determined whether or not (step S406). If the width between the first line element and the white line candidate on the opposite side does not correspond to the road width (No at Step S406), the detection unit 120 proceeds to Step S404.

  On the other hand, when the width between the first line element and the white line candidate on the opposite side corresponds to the road width (Yes at step S406), the detection unit 120 determines that the first line element is a white line candidate (step S407). ), Complete the validity confirmation process.

  Next, processing of the entire LDW device 100 will be described. FIG. 17 is a flowchart showing the processing procedure of the LDW device. As shown in FIG. 17, the LDW device 100 acquires image information (step S10). The LDW device 100 executes a white line candidate detection process (step S11). The white line candidate detection process of step S11 corresponds to the process shown in FIG.

  The departure determining unit 130 of the LDW device 100 determines whether the vehicle is likely to deviate from the white line (step S12). The alarm unit 140 of the LDW device 100 performs alarm notification (step S13) when it is determined by the deviation determination unit 130 that the vehicle is about to deviate from the white line (step S12, Yes). On the other hand, when the departure determining unit 130 does not determine that the vehicle is about to depart from the white line (step S12, No), the LDW device 100 ends the process.

  Next, the effects of the LDW device 100 according to the present embodiment will be described. In the LDW device 100, there is a second line element which is different from the first line element in the traveling direction of the vehicle and has a different direction and intersects the first line element, and the luminance feature of the first line element and the second line element When the luminance feature is the same, it is determined that the first line element does not correspond to the white line candidate. On the other hand, if the second line element does not exist, or if the luminance features of the first and second line elements are not the same, the LDW device 100 determines that the first line element is a white line candidate. Authorized. For this reason, it is possible to extract only white lines such as the lane markings from the various road markings painted on the road without misjudgment by simple processing.

  According to the LDW apparatus 100, the common feature confirmation process does not need to hold a plurality of road surface paint information as dictionary data in advance as in the prior art 2 etc., and confirms only the presence or absence of simple features common to road surface paint in general. Therefore, it can respond to various road surface paint. In addition, it is a simple confirmation process using information obtained by scanning the line element peripheral area that follows the traveling direction once, instead of performing the matching process for the number of types of road surface paint that you want to respond as in the prior art 2 etc. Therefore, the operation cost can be reduced as compared with the prior art 2 and the like. Further, according to the LDW device 100, by performing the validity confirmation processing and the validity confirmation processing, it is possible to cope with a multiple track (decelerating road marking road) that is desired to be detected as a lane boundary.

  According to the LDW device 100, it is further determined whether the width of the region including the first line element and the second line element is smaller than the predetermined area width, and the first line element is determined based on the determination result. It is determined whether is a white line candidate. For this reason, for example, white line candidates can be identified by differentiating with multiple lines and road paint more accurately.

  According to the LDW device 100, it is further determined whether or not the distance between the first line element and the vehicle accompanying the time change is constant, and based on the determination result, whether or not the first line element is a white line candidate Determine if Therefore, it is possible to prevent the traveling direction indication from being erroneously recognized as a white line candidate.

  According to the LDW apparatus 100, it is further determined whether or not there is a white line candidate that makes a pair with the first line element in the image information, and based on the determination result, whether the first line element is a white line candidate Determine if Therefore, the detection accuracy of the white line candidate can be improved by a simple check as to whether or not the width between the white lines corresponds to the road width.

  By the way, in the region including the first line element and the second line element, the detection unit 120 of the LDW device 100 performs the first line element when the area width W is less than the predetermined threshold value Wth in the validation process. Was estimated to be part of a multiple route, but is not limited to this. For example, when the length L in the traveling direction of the region including the first line element and the second line element is equal to or greater than a predetermined threshold Lth, the detection unit 120 sets the first line element to a part of the multiple line. It may be estimated that

  Next, an example of a computer that executes a detection program that implements the same function as the LDW device 100 described in the above embodiment will be described. FIG. 18 is a diagram illustrating an example of a computer that executes a detection program.

  As shown in FIG. 18, the computer 200 has a CPU 201 that executes various types of arithmetic processing, an input device 202 that receives input of data from a user, and a display 203. The computer 200 further includes a reading device 204 that reads a program or the like from a storage medium, and an interface device 205 that exchanges data with other computers via a network. The computer 200 also has a RAM 206 for temporarily storing various information, and a hard disk drive 207. The devices 201 to 207 are connected to the bus 208.

  The hard disk drive 207 has an extraction program 207a and a detection program 207b. The CPU 201 reads the extraction program 207 a and the detection program 207 b and develops the read program on the RAM 206.

  The extraction program 207a functions as an extraction process 206a. The detection program 207b functions as a detection process 206b. The processing of the extraction process 206a corresponds to the processing of the extraction unit 110. The process of the detection process 206 b corresponds to the process of the detection unit 120.

  The extraction program 207a and the detection program 207b may not necessarily be stored in the hard disk drive 207 from the beginning. For example, each program is stored in a "portable physical medium" such as a flexible disk (FD), a CD-ROM, a DVD disk, a magneto-optical disk, an IC card or the like inserted into the computer 200. Then, the computer 200 may read and execute the programs 207a and 207b.

  The following appendices will be further disclosed regarding the embodiment including the above-described respective examples.

(Additional remark 1) The extraction part which extracts from image information the 1st line element which follows in the advancing direction of vehicles as a white line candidate,
In the image information, when there is a second line element which is different from the first line element and intersects the first line element, the luminance feature of the first line element and the second line element And detecting the white line candidate when the luminance features of the first line element and the second line element are the same, and detecting a white line. Detection device.

(Supplementary Note 2) The detection unit further determines whether the width of the region including the first line element and the second line element is wider than a predetermined area width, and the first line element and the second line element The detection device according to claim 1, wherein the white line is detected by excluding the white line candidate when the width of the area including the second line element is wider than a predetermined area width.

(Supplementary Note 3) The detection unit further determines whether the length in the traveling direction of the region including the first line element and the second line element is shorter than a predetermined length, and the first line The detection device according to Supplementary Note 1, wherein the white line is detected by excluding it from the white line candidate when the length in the traveling direction of the region including the second line element and the second line element is shorter than a predetermined length. .

(Supplementary Note 4) The detection unit further determines whether the distance between the first wire element and the vehicle with time change is constant, and the first wire element and the vehicle with time change. The detection device according to Supplementary Note 1, wherein the white line is detected by excluding the white line candidate when the distance is not constant.

(Supplementary Note 5) The detection unit further determines whether or not there is another first line element paired with the first line element in the image information, and further forms a pair with the first line element The detection device according to claim 1, wherein the white line candidate is excluded from the white line candidate and the white line is detected when the first line element of the first line element does not exist.

(Supplementary Note 6) A detection method executed by a computer,
The first line element continuous in the traveling direction of the vehicle is extracted from the image information as a white line candidate,
In the image information, when there is a second line element which is different from the first line element and intersects the first line element, the luminance feature of the first line element and the second line element To determine if is the same,
A detection method characterized by performing processing of detecting a white line by excluding it from the white line candidate when the luminance features of the first line element and the second line element are the same.

(Supplementary Note 7) The determination process further determines whether the width of the region including the first line element and the second line element is wider than a predetermined area width, and the first line element The detection method according to claim 6, wherein the white line is detected by excluding the white line candidate when the width of the area including the second line element is wider than a predetermined area width.

(Supplementary Note 8) The determination process further determines whether the length in the traveling direction of the region including the first line element and the second line element is shorter than a predetermined length, The detection according to claim 6, wherein the white line is detected and the white line is detected when the length of the region including the line element and the second line element in the traveling direction is shorter than a predetermined length. Method.

(Supplementary Note 9) The determination process further determines whether the distance between the first wire element and the vehicle according to the time change is constant, and the first wire element and the vehicle according to the time change. The detection method according to claim 6, wherein the white line candidate is excluded from the white line candidate and the white line is detected when the distance is not constant.

(Supplementary Note 10) The determination process further determines whether or not there is another first line element paired with the first line element in the image information, and becomes a pair with the first line element The detection method according to claim 6, wherein the white line candidate is excluded from the white line candidate and the white line is detected when no other first line element is present.

(Supplementary note 11)
The first line element continuous in the traveling direction of the vehicle is extracted from the image information as a white line candidate,
In the image information, when there is a second line element which is different from the first line element and intersects the first line element, the luminance feature of the first line element and the second line element To determine if is the same,
A detection program characterized by excluding processing from the white line candidate and detecting a white line when the luminance features of the first line element and the second line element are the same.

(Supplementary Note 12) The determination process further determines whether the width of the region including the first line element and the second line element is wider than a predetermined area width, and the first line element The detection program according to claim 11, wherein the white line is detected by excluding the white line candidate when the width of the area including the second line element is wider than a predetermined area width.

(Supplementary note 13) The determination process further determines whether the length in the traveling direction of the region including the first line element and the second line element is shorter than a predetermined length, The white line is detected by excluding it from the white line candidate when the length in the traveling direction of the area including the line element and the second line element is shorter than a predetermined length, and the white line is detected. program.

(Supplementary Note 14) The determination process further determines whether the distance between the first wire element and the vehicle with time change is constant, and the first wire element and the vehicle with time change The detection program according to appendix 11, wherein the white line candidate is excluded from the white line candidate and the white line is detected when the distance of is not constant.

(Supplementary Note 15) The determination process further determines whether or not there is another first line element paired with the first line element in the image information, and becomes a pair with the first line element The non-transitory computer-readable storage medium according to claim 11, wherein the white line candidate is excluded from the white line candidate and a white line is detected when no other first line element is present.

100 LDW device 110 extraction unit 120 detection unit 130 deviation judgment unit 140 alarm unit

Claims (7)

An extraction unit which extracts, from the image information, a first line element continuous in the traveling direction of the vehicle as a white line candidate;
In the image information, when there is a second line element which is different from the first line element and intersects the first line element, the luminance feature of the first line element and the second line element And detecting the white line candidate when the luminance features of the first line element and the second line element are the same, and detecting a white line. Detection device.   The detection unit further determines whether the width of the area including the first line element and the second line element is wider than a predetermined area width, and the first line element and the second line element The detection apparatus according to claim 1, wherein the white line is detected by excluding the white line candidate from the white line candidate when the width of the area including W is wider than a predetermined area width.   The detection unit further determines whether the length in the traveling direction of the region including the first line element and the second line element is shorter than a predetermined length, and the first line element and the second line element The detection device according to claim 1, wherein the white line is detected by excluding the white line candidate from the white line candidate when the length in the traveling direction of the region including the two line elements is shorter than a predetermined length.   The detection unit further determines whether the distance between the first wire element and the vehicle with time change is constant, and the distance between the first wire element and the vehicle with time change is not constant. The detection device according to claim 1, wherein the detection unit is excluded from the white line candidate in some cases and detects the white line.   The detection unit further determines whether or not there is another first line element paired with the first line element in the image information, and the other first line paired with the first line element The detection apparatus according to claim 1, wherein the detection unit excludes the white line candidate and detects a white line when no element is present. A computer implemented detection method,
The first line element continuous in the traveling direction of the vehicle is extracted from the image information as a white line candidate,
In the image information, when there is a second line element which is different from the first line element and intersects the first line element, the luminance feature of the first line element and the second line element To determine if is the same,
A detection method characterized by performing processing of detecting a white line by excluding it from the white line candidate when the luminance features of the first line element and the second line element are the same. On the computer
The first line element continuous in the traveling direction of the vehicle is extracted from the image information as a white line candidate,
In the image information, when there is a second line element which is different from the first line element and intersects the first line element, the luminance feature of the first line element and the second line element To determine if is the same,
A detection program characterized by excluding processing from the white line candidate and detecting a white line when the luminance features of the first line element and the second line element are the same.

Images