JP6387795B2 - Illumination light spot detection device, spot detection processing method, program, and storage medium - Google Patents

Description

  The present invention relates to an illumination lamp astigmatism detection apparatus, an astigmatism detection processing method, a program, and a storage medium.

  Patent Literature 1 discloses an illuminating lamp identification device including an image data acquisition unit, a position information acquisition unit, an illuminance distribution calculation unit, a storage unit, a comparison unit, and an identification unit. The image data acquisition unit acquires image data of a plurality of illuminating lamps arranged along the road, and the position information acquisition unit acquires position information of a moving body that travels along the road. The illuminance distribution calculation unit calculates the first illuminance distribution from the relationship between the image data acquired by the image data acquisition unit and the travel distance measured by the position information acquisition unit. The storage unit stores a second illuminance distribution in advance. The comparison unit compares the first illuminance distribution with the second illuminance distribution, and the identification unit performs illumination based on the difference between the first illuminance distribution and the second illuminance distribution at the predetermined travel distance compared by the comparison unit. Identify the illumination status and position of the lamp.

  Patent Document 2 discloses an astigmatic illuminating lamp identification device including a lighting detection unit, a DGPS receiver, a storage unit, and a processing unit. The lighting detection unit is an illuminance meter or imaging means, and is mounted on a vehicle that travels along a route in which a plurality of illumination lamps are arranged, and is illuminated from the illumination lamps when positioned near each of the plurality of illumination lamps. The lighting state of the illuminating lamp is detected based on the light. The DGPS receiver acquires positioning data indicating the position of the vehicle. The storage unit stores illumination lamp position information indicating the positions of the plurality of illumination lamps. Based on the output from the lighting detection unit, the positioning data obtained by the DGPS receiver, and the illuminating lamp position information stored in the storage unit, the processing unit detects that the vehicle is near the illuminating lamp. An illumination lamp whose lighting state is not detected by the lighting detection unit when positioned is identified as a lighting lamp in a non-lighting state.

JP 2004-146281 A JP 2001-85174 A

  However, according to the configuration of Patent Document 1, a position information acquisition unit that acquires travel position information of a moving body is required for detecting the inconspicuousness of an illumination lamp, and the storage unit is configured to have an initial illuminance distribution (second illuminance distribution). ) Must be remembered. Also in the configuration of Patent Document 2, a DGPS receiver that acquires positioning data indicating the vehicle position is required, and the storage unit needs to previously store illumination lamp position information indicating the position of the illumination lamp. Therefore, in the above-described configurations of Patent Documents 1 and 2, a positioning device for acquiring and inputting position information of a moving body such as a vehicle is required, so that the astigmatism detection device is complicated and expensive. There's a problem. In addition, since the initial illumination lamp position information is managed by the equipment designer, it may be difficult for a worker who detects the incongruity to be difficult to obtain. There has been a problem that it is difficult to detect astigmatism.

  Then, this invention makes it a subject to provide the simple and simple astigmatism detection apparatus and the astigmatism detection method which do not require the input of moving body position information and initial illumination light position information regarding the astigmatism detection of an illumination lamp. To do.

  An astigmatism detection device for detecting an illuminance of an illuminating lamp according to the present invention includes an image acquisition unit that acquires a moving image including a plurality of illuminating lamps photographed during traveling of a moving body, and a detection in a display screen that displays the moving image. And an image analysis unit that identifies an illuminating lamp out of the plurality of lamps based on the periodicity of spots corresponding to the plurality of lamps passing through the region.

  According to the above astigmatism detection device, the image analysis unit is configured to identify the astigmatic illumination lamp based on the periodicity of the spot passing through the detection area in the moving image display screen. At the time of point detection, the configuration for acquiring the moving body position information and the acquisition and input of the initial illumination lamp position information are not required, and a simple and simple point detection device is realized.

  In the first form of the astigmatism detection device, image analysis is performed when the moving body travels at a substantially constant speed and the first to nth spots corresponding to the plurality of illumination lights pass through the detection region in this order. A region setting unit for setting a detection region, an interval calculation unit for calculating an average time interval in which the first to xth spots pass through the detection region, and an average time interval for spots after x + 1 And a prediction determination unit that determines whether or not to appear in the detection area every time, and a disadvantage that specifies a spotlight that is not lit based on the fact that the spot after the (x + 1) th time does not appear in the detection area at every average time interval A specific part. Thus, since the presence or absence of a defect is determined in a specific detection region and timing, it is possible to reliably identify the defect with a relatively small amount of data processing.

  Here, the detection area is set at the upper right end or upper left upper end of the display screen, and the count value m × the average time interval has elapsed from the time when the x-th spot has passed through the detection area. When there is no spot, the astigmatism specifying unit is configured to specify that the x + m illumination lamp is astigmatism. As a result, it is possible to reliably identify the position of the illuminating lamp with a simple processing configuration.

  In the second form of the astigmatism detection device, image analysis is performed when the moving body travels at a substantially constant speed and the first to n-th spots corresponding to a plurality of illumination lights pass through the detection region in this order. A region setting unit for setting the first and second detection regions as detection regions, and an interval calculation unit for calculating an average time interval in which the first to xth spots pass through the first detection region; The prediction determination unit that determines whether or not the spot after x + 1 appears in the second detection area at every average time interval, and the second detection at the spot after x + 1 at every average time interval An image portion that includes a non-spot identifying unit that identifies a non-spot illuminating lamp based on the fact that it does not appear in the area, and at least the second detection area indicates an installation interval between the first illuminating lamp and the second illuminating lamp. It is prescribed by. Thereby, since the presence or absence of the spot is determined using the second detection area having a relatively wide width, even if the detection timing by the prediction detection unit is slightly deviated, the spot or the position that is supposed to be the spot (In other words, the position corresponding to the illuminating lamp) is surely included in the second detection region. In addition, since the width of the second detection area is a length that does not include two or more spots, even if the detection timing by the prediction detection unit is slightly shifted, the detection timing order and the spot or spot to be detected There is no discrepancy in the order of positions that should be. Therefore, it is possible to reliably determine the inconvenience for each of the illumination lamps.

  Here, when the x + m-th spot does not exist in the second detection area at the time when the count value m × the average time interval has elapsed from the time when the x-th spot has passed through the first detection area, the astigmatism specifying unit Is configured to identify that the x + m illuminating lamp is astigmatic. As a result, it is possible to reliably identify the position of the illuminating lamp with a simple processing configuration.

  Further, in the above astigmatism detection device, the moving body travels at a substantially constant speed, and the image analysis unit calculates a time interval for the spot to pass through the detection region, and a specific calculation calculated by the interval calculation unit. When the time interval is a multiple of another time interval, an incongruity specifying unit that specifies that the illumination lamp corresponding to the period included in the specific time interval is inconsequential may be provided. This further simplifies the configuration of the astigmatism detection device.

  The present invention further includes a program for causing a computer to function as each unit included in the above-described defect detection device. The present invention also includes a computer-readable storage medium storing the program. Furthermore, the present invention also includes an astigmatism detection method including a step of executing processing of each part of the astigmatism detection device.

It is the schematic for demonstrating use of the non-point detection apparatus by this invention. It is a figure which shows an example of 1 frame of the moving image used with the astigmatism detection apparatus by the 1st Embodiment of this invention. It is a block diagram of the astigmatism detection apparatus of a 1st embodiment. It is a figure explaining operation | movement of the astigmatism detection apparatus by 1st Embodiment. It is a flowchart of the point detection method of a 1st embodiment. It is a figure which shows an example of one frame of the moving image used with the astigmatism detection apparatus by the 2nd Embodiment of this invention. It is a figure which shows the other example of 1 frame of the moving image used with the astigmatism detection apparatus by the 2nd Embodiment of this invention. It is a flowchart of the defect detection method of 2nd Embodiment.

<Outline configuration>
FIG. 1 shows a schematic diagram of the use of a defect detection device 1 according to the present invention. A vehicle V, which is a moving body, travels on the road surface R, and the in-vehicle camera C mounted on the vehicle V causes the illumination lights 2-1 to 2-n above the right lane and the illumination lights 3-1 to 3- above the left lane. n 'videos are taken. In each of the following embodiments, the lighting lamps 2-1 to 2-n and 3-1 to 3-n ′ are tunnel lights installed from the ceiling to the side wall in the tunnel T, and the vehicle V has a wellhead Ta (hereinafter, It is assumed that the vehicle travels from the “entrance Ta” toward the wellhead Tb (hereinafter referred to as “exit Tb”). Further, in the following description, the lighting lamps 2-1 to 2-n are collectively referred to as the lighting lamp 2 as a representative or a part thereof, and the lighting lamp 2 is from the entrance Ta side to the exit Tb side. It is assumed that the lighting lamps 2-1 to 2-n are arranged in this order. Similarly, the illuminating lamps 3-1 to 3-n ′ are collectively referred to as the illuminating lamp 3 as a representative or a part thereof, and the illuminating lamp 3 is directed from the entrance Ta side to the exit Tb side. The lighting lamps 3-1 to 3-n ′ are arranged in this order.

  The astigmatism detection device 1 analyzes the moving image M taken by the in-vehicle camera C, detects whether or not the illuminating lamp 2 or 3 includes an astigmatic illuminating lamp, and further, which illuminating lamp 2 or 3 It is determined whether or not is a defect. The astigmatism detection device 1 may be mounted or disposed on the vehicle V or may be disposed at a location different from the vehicle V. As shown in FIG. 1, the moving image M is displayed on a display screen 31 defined by a right end 31a, a left end 31b, an upper end 31c, and a lower end 31d. When the in-vehicle camera C captures the front of the vehicle V in the traveling direction, on the display screen 31, spots Sr1 to Srn corresponding to the illumination lights 2-1 to 2-n are located at the right end 31a from the center of the screen. It moves toward the vicinity of the upper end 31c or the right end 31a of the upper end 31c (hereinafter referred to as “upper right end”). In addition, spots Su1 to Sun ′ that are bright portions corresponding to the illumination lamps 3-1 to 3-n ′ are displayed on the display screen 31 from the center of the screen, near the upper end 31c of the left end 31b or near the left end 31b of the upper end 31c ( Hereinafter, it moves toward “the upper left end”.

  In each figure, the moving image M is not shaded for clarity of explanation, but it is understood that the actual moving image M has the following shades based on a general tunnel configuration. The The image portions corresponding to the illumination lamps 2 and 3 (that is, the spots Sr1 to Srn and Su1 to Sun ′) are the brightest, and the image portions corresponding to the tunnel ceiling Tc and the portions excluding the spots are the darkest. Then, the image portion corresponding to the tunnel ceiling Tc becomes brighter from the image portion corresponding to the tunnel side wall Ts, and the image portion corresponding to the tunnel side wall Ts and the image portion corresponding to the road surface R have the same brightness.

  In the present invention, since the illumination lamps 2-1 to 2-n and 3-1 to 3-n ′ are arranged at substantially equal intervals, the illumination lamps 2-1 to 2-n and 3-1 to n are arranged. Note that if 'is lit, the corresponding spots Sr1 to Srn and Su1 to Sun' are also displayed on the moving image M in a periodic manner. In other words, the present invention identifies the disadvantage based on this periodicity. In the following description, the spots Sr1 to Srn are collectively referred to as a spot Sr as a representative of some of them, and the spots Su1 to Sun ′ are collectively referred to as a spot or a part thereof. As a representative, it is assumed to be a spot Su. Hereinafter, the illumination lamp 2 and the spot Sr will be described.

<First Embodiment>
FIG. 2 shows one frame of the moving image M used in the astigmatism detection device 1 according to the first embodiment. As an outline, the astigmatism detection device 1 identifies the presence and position of the astigmatic illumination lamp 2 by detecting the spot Sr passing through the detection area 32 at the upper right end of the display screen 31. The astigmatism detection device 1 obtains an average time interval Δt at which the spots Sr corresponding to several illumination lamps 2 on the entrance Ta side pass through the detection area 32 and predicts the timing at which the subsequent spot Sr appears in the detection area 32. Then, based on whether or not the spot Sr is detected at the predicted timing, it is determined whether each illumination lamp 2 is turned on or not.

  FIG. 3 shows a block diagram of the astigmatism detection device 1 of the present embodiment. The astigmatism detection device 1 includes an image acquisition unit 10, an image analysis unit 20, an output unit 30, and an operation input unit 35. The image acquisition unit 10 is an input interface for taking in the moving image M captured by the in-vehicle camera C into the astigmatism detection device 1. In the image acquisition unit 10, the moving image M may be acquired from the in-vehicle camera C via a wired connection or a wireless connection, or via a buffer (memory) on the in-vehicle camera C, or recorded from the in-vehicle camera C. The moving image M may be acquired via a medium. In the former case, the image acquisition unit 10 functions as a reception unit. The image acquisition unit 10 inputs the acquired moving image M to the image analysis unit 20.

  The image analysis unit 20 includes a CPU 21 and a memory 26, and the CPU 21 includes an area setting unit 22, an interval calculation unit 23, a prediction determination unit 24, and an astigmatism specifying unit 25. These units are connected to each other through a bus 29 so that they can communicate with each other, and in the CPU 21, the exchange of signals between the units is appropriately controlled. The CPU 21 also includes general-purpose functions such as a timer, a counter, and moving image operating means such as moving image reproduction. The memory 26 is a memory such as a RAM or ROM that stores programs and data.

  The area setting unit 22 is for setting the position, shape, size, and the like of the detection area 32. In the present embodiment, the detection area 32 may be set in advance from the general size of the illuminating lamp, the general height with respect to the road surface R, and the like, and the spot Sr related to the illuminating lamp 2 may be set in advance. It is set at the upper right corner of the screen (the spot Su related to the illumination lamp 3 is set at the upper left corner of the screen). Alternatively, the detection area 32 may be set by a user operation. For example, the region setting unit 22 may determine the detection region 32 based on a user input performed via an operation input unit 35 described later (for example, by a mouse operation or the like).

  The interval calculator 23 calculates an average value Δt of time intervals during which the spots Sr1 to Srx pass through the detection area 32 of the display screen 31. Specifically, the interval calculation unit 23 performs the time tk when the spot Srk passes the detection region 32 and the time t (k) when the spot Sr (k−1) passes the detection region 32 for k of 2 ≦ k ≦ x. −1) is calculated as a time interval Δsk = tk−t (k−1), and the total of Δs2 to Δsx is divided by x−1 to calculate an average time interval Δt. x is a predetermined value, preferably about x = 4-6.

  Here, the calculation of the average time interval Δt will be described with reference to FIG. FIG. 4 is a schematic diagram of the display screen 31 when the spots Sr <b> 1 to Sr <b> 5 pass through the detection region 32. In FIG. 4, a spot Sr corresponding to five illumination lamps 2 from the entrance Ta side is shown. In this example, it is assumed that x = 5, and the illumination lights 2-1 to 2-5 are normally lit. Further, the vehicle V is assumed to travel at a substantially constant speed during shooting of the moving image M, and the time display 34 indicates an elapsed time from the start of shooting. As will be described in detail later, the spot Sr8 does not exist.

  As shown in (a), the interval calculation unit 23 acquires time t1 (36 seconds 20 in this example) when the spot Sr1 passes through the detection region 32, and stores the time t1 in the memory 26. The time when each spot Sr passes through the detection area 32 may be the time when the luminance detected in the detection area 32 becomes maximum, or the time when the detected luminance exceeds a predetermined threshold. Alternatively, it may be the time of the moment when the detected luminance exceeds the predetermined threshold and then becomes less than the threshold. Thereafter, the moving image M is advanced to a state (b).

As shown in (b), the interval calculation unit 23 obtains a time t2 (in this example, 37 seconds 08) when the spot Sr2 passes the detection region 32, and is a time interval that is a difference between the time t2 and the time t1. Δs2 = t2−t1 (= 0.18 seconds) is calculated, and the time interval Δs2 and time t2 are stored in the memory 26. Thereafter, the moving image M is advanced to enter the state (c).
As shown in (c), as in (b), the interval calculation unit 23 acquires time t3 (in this example, 37 seconds 26) when the spot Sr3 passes through the detection region 32, and time t3 and time The time interval Δs3 = t3-t2 (= 0.18 seconds), which is the difference between t2, is calculated, and the time interval Δs3 and time t3 are stored in the memory 26. Thereafter, the moving image M is advanced to enter the state (d).
As shown in (d), similarly to (b), the interval calculation unit 23 acquires time t4 (in this example, 38 seconds 14) when the spot Sr4 passes through the detection region 32, and time t4 and time The time interval Δs4 = t4-t3 (= 0.18 seconds), which is the difference between t3, is calculated, and the time interval Δs4 and time t4 are stored in the memory 26. Thereafter, the moving image M is advanced to a state (e).
As shown in (e), as in (b), the interval calculator 23 obtains the time t5 (in this example, 39 seconds 01) when the spot Sr5 passes through the detection region 32, and the time t5 and time The time interval Δs5 = t5-t4 (= 0.17 seconds), which is the difference between t4, is calculated, and the time interval Δs5 is stored in the memory 26.

  The interval calculation unit 23 calculates an average value of Δs2 (0.18 seconds), Δs3 (0.18 seconds), Δs4 (0.18 seconds) and Δs5 (0.17 seconds) stored in the memory 26. Obtain the average time interval Δt. In this example, Δt = 0.1775 seconds. Note that the process of advancing the moving image M may be performed by a user operation via the operation input unit 35.

  The prediction determination unit 24 determines whether or not the spot Sr appears in the detection region 32 at each time that is a multiple of the average time interval Δt after time tx (in this example, time t5). Specifically, the prediction determination unit 24 determines whether or not the spot Sr (x + m) exists in the detection region 32 at the m-th prediction timing (time tx + m × Δt) after time tx. The prediction determination unit 24 stores the count value m and the determination value d indicating the presence / absence of the spot Sr in the memory 26. In the present embodiment, it is assumed that the discriminant value d is d = 1 when there is a spot and d = 0 when there is no spot. As described above, the prediction determination unit 24 shifts the moving image M by the average time interval Δt from the state in which the detection area 32 includes the spot Srx, and whether or not the detection area 32 includes the spot Sr at each detection timing. Is determined.

  Regarding the determination of the presence / absence of the spot Sr in the detection area 32, the prediction determination unit 24 has the spot Sr when the maximum luminance in the detection area 32 exceeds a predetermined threshold, and does not have the spot Sr otherwise. Can be determined. In addition, the prediction detection unit 24 may determine that the spot Sr exists when the average luminance in the detection region 32 exceeds a predetermined threshold, and otherwise does not include the spot Sr.

  When the prediction determination process ends, the prediction determination unit 24 outputs an end signal to the astigmatism specifying unit 25. Even if the total number n of the illuminating lamps 2 is unknown, it can be identified by various methods that a certain spot Sr is the last spot. For example, after a certain spot Sr appears in the detection area 32, another spot does not appear in the detection area 32 for a sufficiently long predetermined period (that is, for a considerable number of m) (in the case of nighttime). It may be possible to identify that the spot Sr is the last spot by, for example, a sharp increase in brightness (in the case of daytime). Alternatively, end information indicating the end of shooting (for example, a stop image at the end of shooting) may be added to the moving image M, and the last spot may be identified based on the end information. Alternatively, the last spot may be identified by visual observation of the display screen 31 by the user (for example, visual observation that the moving image M has exited the exit Tb).

  When the end signal is input, the astigmatism specifying unit 25 reads the combination of the count value m and the determination value d stored in the memory 26. When the discriminant value d = 1 for all the count values m, the astigmatism specifying unit 25 indicates that all of the illumination lamps 2-1 to 2-n are lit (or there is no astigmatism illumination lamp 2). All the lighting information indicating that is output to the output unit 30. On the other hand, when the discriminant value d = 0 for the specific count value m = y, the astigmatism specifying unit 25 provides the output unit 30 with astigmatism information indicating that the illumination lamp 2- (x + y) is astigmatism. Output. The output of all lighting information and astigmatism information by the output unit 30 may be a display output on the display screen 31, a voice output by voice guidance or the like, or these display outputs and voice outputs It may be a combination.

  For example, assuming that the illumination lamp 2-8 is inconsequential among the illumination lamps 2-1 to 2-10 (see FIGS. 4D and 4E), the process of the prediction determination unit 24 is as follows. It becomes like this. The prediction determination unit 24 determines whether or not the spot Sr6 exists in the detection region 32 at the prediction timing (t5 + 1 × Δt) when the count value m = 1. Since the spot Sr6 exists, the prediction determination unit 24 causes the memory 26 to store a combination of the count value m = 1 and the determination value d = 1. Similarly, the prediction determination unit 24 determines whether or not the spot Sr7 exists in the detection region 32 at the prediction timing (t5 + 2 × Δt) of the count value m = 2. Since the spot Sr7 exists, the prediction determination unit 24 stores the combination of the count value m = 2 and the determination value d = 1 in the memory 26. The prediction determination unit 24 determines whether or not the spot Sr8 exists in the detection region 32 at the prediction timing (time t5 + 3 × Δt) of the count value m = 3. Since the spot Sr8 does not exist, the prediction determination unit 24 causes the memory 26 to store a combination of the count value m = 3 and the determination value d = 0. For the count values m = 4 and m = 5, similarly to the case of m = 1, the prediction determination unit 24 determines that the count value m = 4 and the determination value d for the prediction timing (t5 + 4 × Δt) of the count value m = 4. = 1, the combination of the count value m = 5 and the discriminant value d = 1 is stored in the memory 26 with respect to the prediction timing (t5 + 5 × Δt) of the count value m = 5. When receiving the end signal, the astigmatism specifying unit 25, based on the combination of the count value m = 3 (y = 3) and the discrimination value d = 0 stored in the memory 26, the illumination lamp 2- (5 + 3), that is, Astigmatism information indicating that the illuminating lamp 2-8 is astigmatism is output to the output unit 30.

  The output unit 30 includes a monitor having a display screen 31 and a necessary output interface. The monitor may be a monitor such as a desktop, a notebook computer, or a tablet, and may be a monitor dedicated to the astigmatism detection device 1 or a general-purpose monitor. The operation input unit 35 is a mouse, a keyboard, a touch panel integrated with the display screen 31, and the like, and is an input interface that receives a user operation. As described above, the operation input unit 35 can be used in the setting of the detection region 32 by the region setting unit 22, the operation of the moving image M in the acquisition of the time interval Δs by the interval calculation unit 23, and the like.

FIG. 5 shows an example of a flowchart of the defect detection method according to the present embodiment.
In step S <b> 100, the vehicle V travels through the tunnel T, and the moving image M including the illuminating lamp 2 is captured by the in-vehicle camera C.
In step S <b> 105, the image acquisition unit 10 inputs the moving image M acquired from the in-vehicle camera C to the image analysis unit 20.

  In step S110 and subsequent steps, image analysis processing by the image analysis unit 20 is performed. In step S110, the area setting unit 22 sets a predetermined detection area 32 or sets the detection area 32 based on a user input. The count values k and m for the spot Sr are set to k = 1 and m = 0.

In step S <b> 120, the interval calculation unit 23 acquires time t <b> 1 when the spot Sr <b> 1 passes through the detection region 32, and stores time t <b> 1 in the memory 26.
In step S122, the interval calculation unit 23 increments the count value k of the spot Sr by 1, and further advances the moving image M.
In step S124, the interval calculation unit 23 obtains a time tk when the spot Srk passes through the detection region 32, calculates a time interval Δsk that is a difference between the time tk and the time t (k−1), and stores the time in the memory 26. The interval Δsk and time tk are stored.

In step S126, the interval calculation unit 23 determines whether or not the count value k is k = x. If k <x (step S126: NO), the process returns to step S122. If k = x (step S126: YES), the process proceeds to step S128. That is, at the end of step S126, the memory 26 stores the time intervals Δs2 to Δsx.
In step S128, the interval calculator 23 calculates the average time interval Δt by dividing the sum of the time intervals Δs stored in the memory 26 by x-1.

In step S130, the prediction determination unit 24 advances the moving image M by the average time interval Δt and increments the count value m by 1.
In step S132, the prediction determination unit 24 determines whether or not the spot Sr (x + m) exists in the detection region 32 at the time tx + m × Δt. When the spot Sr (x + m) exists (step S132: YES), in step S134, the prediction determination unit 24 stores the count value m and the determination value d = 1 in the memory 26. On the other hand, when the spot Sr (x + m) does not exist (step S132: NO), in step S136, the prediction determination unit 24 stores the count value m and the determination value d = 0 in the memory 26.
In step S138, the prediction determination unit 24 determines whether or not the prediction determination for the last spot Sr has been completed. If the prediction determination has not ended (step S138: NO), the process returns to step S130. When the prediction determination ends (step S138: YES), the prediction determination unit 24 outputs an end signal, and the process proceeds to step S140.

  In step S140, the astigmatism specifying unit 25 determines whether or not the discriminant value d = 1 for all count values m. When the discriminant value d = 1 for all count values m (step S140: YES), the process proceeds to step S150. On the other hand, when the discriminant value d = 0 for the specific count value m (m = y) (step S140: NO), the process proceeds to step S155.

In step S150, the astigmatism specifying unit 25 causes the output unit 30 to output all lighting information indicating that all of the illuminating lamps 2 are lit (or that there is no illuminating illuminating lamp 2).
In step S155, the astigmatism specifying unit 25 causes the output unit 30 to output astigmatism information indicating that the illuminating lamp 2- (x + y) is an astigmatism based on the count value y. As a result, the spotlight 2 is detected and specified.

  As described above, according to the astigmatism detection device 1 of the present embodiment, the image analysis unit 20 is based on the periodicity of the spot Sr that passes through the detection region 32 in the display screen 31 that displays the acquired moving image M. Identify unlit lighting. This eliminates the need for a structure for acquiring moving body position information and the acquisition and input of the initial illumination light position information when detecting the astigmatism of the illuminating lamp, thereby realizing a simple and simple astigmatism detecting device and astigmatism detecting method. Is done.

  In the image analysis unit 20, the interval calculation unit 23 calculates the average time interval Δt in which the spots Sr1 to Srx pass through the detection region 32, and the prediction determination unit 24 averages the spots Sr after the spot Sr (x + 1). It is determined whether or not it appears in the detection area 31 at every time interval Δt, and if it does not appear, the astigmatism specifying unit 25 specifies that an astigmatic illuminating lamp exists. Thereby, since the presence or absence of a defect is determined in a specific detection region and timing, it becomes possible to reliably identify the defect with a relatively small amount of data processing.

  Further, when the detection area 32 is set at the upper right end of the display screen 31 and the spot Sr (x + m) does not exist in the detection area 32 at the time when m × Δt has elapsed from the time when the spot Srk has passed through the detection area 32. In addition, the astigmatism specifying unit 25 is configured to specify that the illuminating lamp 2- (x + m) is an astigmatism. As a result, it is possible to reliably identify the position of the illuminating lamp with a simple processing configuration.

<Second Embodiment>
In the first embodiment, the configuration in which the detection region 32 is used in the calculation of the average time interval Δt and the subsequent prediction detection is shown. However, in the present embodiment, in the calculation of the average time interval Δt and the subsequent prediction detection. Fig. 3 shows a configuration in which different detection areas are used. Note that a detection area used in predictive detection is referred to as a mask area 33.

  FIG. 6 shows one frame of the moving image M that is used particularly in the prediction detection process of the astigmatism detection apparatus 1 of the present embodiment. As shown in FIG. 6, the mask region 33 is determined based on the positional relationship between the illumination lamp 2-1 and the illumination lamp 2-2, that is, the positional relationship on the screen between the spots Sr1 and Sr2. Specifically, the mask area 33 is a quadrangle formed of a right side 33a, a left side 33b, an upper side 33c, and a bottom side 33d, and the right side 33a and the left side 33b are parallel to each other. The right side 33a and the left side 33b are set corresponding to the perpendicular of the electrical wiring W connecting the illumination lamp 2-1 and the illumination lamp 2-2. Note that the right side 33a and the left side 33b may be set to be perpendicular to an auxiliary line (not shown) connecting the spots Sr1 and Sr2. The right side 33a and the left side 33b may have the same length or different lengths, but the right side 33a and the left side 33b are set to intersect at least the wiring W or the auxiliary line. Note that the intersection 33e between the right side 33a and the upper side 33c or the intersection 33f between the right side 33a and the bottom side 33d may not necessarily be in the display screen 31 (in this case, the mask region 33 is a pentagon or a hexagon).

  The separation distance between the right side 33a and the left side 33b is the distance between the entrance Ta side end of the spot Sr1 and the entrance Ta side end of the spot Sr2 (an image portion showing the installation interval between the illumination lamp 2-1 and the illumination lamp 2-2). Similarly, it may be set to the distance between the exit Tb side end of the spot Sr1 and the exit Tb side end of the spot Sr2. Further, the separation distance between the right side 33a and the left side 33b may be set to the distance between the exit Tb side end of the spot Sr1 and the entrance Ta side end of the spot Sr2 (that is, the separation distance between the spot Sr1 and the spot Sr2). In this embodiment, the separation distance between the right side 33a and the left side 33b is set to the distance between the entrance Ta side end of the spot Sr1 and the entrance Ta side end of the spot Sr2.

  In FIG. 6, the upper side 33c and the bottom side 33d are set substantially parallel to the wiring W or the auxiliary line, and the mask region 33 is generally rectangular as a whole. However, as shown in FIG. 31 may be set on the upper end 31 c of the display 31, or the base 33 d may be set on the right end 31 a of the display screen 31. In the configuration using the substantially rectangular mask area 33 as shown in FIG. 6, the area to be subjected to image processing is relatively small, and the data processing amount is reduced. In addition, the possibility that other light sources (such as guide lights) that can appear below the display screen 31 are included in the mask region 33 is reduced, and the accuracy of the detection process is improved. On the other hand, in the configuration using the trapezoidal or belt-like mask region 33 as shown in FIG. 7, if the right side 33a and the left side 33b are determined (or even the intersection of the right side 33a and the left side 33b and the wiring W or the auxiliary line is determined. If the upper side 33c and the lower side 33d are automatically set, the data processing amount for the setting process of the mask area 33 is reduced.

  The block diagram of the astigmatism detection device 1 of the present embodiment is the same as the block diagram of the astigmatism detection device 1 of the first embodiment shown in FIG. However, in the first embodiment, only the given detection area 32 is set by the area setting unit 22, whereas in the present embodiment, the above-described mask area is set together with the setting of the detection area 32 by the area setting unit 22. 33 is set.

  The area setting unit 22 sets the detection area 32 at the upper right end of the display screen 31 as in the first embodiment. Thereby, the interval calculation unit 23 calculates the average time interval Δt, as in the first embodiment. Then, the area setting unit 22 sets the mask area 33 based on the spots Sr1 and Sr2. Similar to the setting of the detection area 32, the setting of the mask area 33 may be performed by user input. In this case, the region setting unit 22 sets the mask region 33 based on a user operation via the operation input unit 35. Note that it is possible that only the mask region 33 is set by the region setting unit 22 without setting the detection region 32, and the interval calculation unit 23 calculates the average time interval Δt using the mask region 33.

  The prediction determination unit 24 determines whether or not the spot Sr appears in the mask region 33 at each time that is a multiple of the average time interval Δt after the time tx (in this example, time t5). Specifically, the prediction determination unit 24 determines whether or not a spot Sr (x + m) exists in the mask region 33 at the m-th prediction timing (time tx + m × Δt) after time tx. The prediction determination unit 24 stores the count value m and a determination value d indicating the presence / absence of the spot Sr (d = 1 when there is a spot and d = 0 when there is no spot) in the memory 26. That is, the prediction determination unit 24 shifts the moving image M by the average time interval Δt from the state in which the mask area 33 includes the spot Sr, and determines whether the mask area 33 includes the spot Sr at each detection timing. Determine. Similar to the first embodiment, the astigmatism specifying unit 25 receives the end signal from the prediction determination unit 24, and based on the count value m and the discrimination value d stored in the memory 26, the all lighting information or the astigmatism. The information is output to the output unit 30.

FIG. 8 shows an example of a flowchart of the defect detection method according to the present embodiment.
The processing in steps S100 to S110 is the same as the processing in the first embodiment shown in FIG. That is, in step S100, the moving image M is photographed, and in step S105, the image acquisition unit 10 inputs the moving image M to the image analysis unit 20.

In step S110 and subsequent steps, image analysis processing by the image analysis unit 20 is performed.
In step S <b> 110, the detection area 32 is set by the area setting unit 22.
In step S115, the mask region 33 is set by the region setting unit 22 based on the spots Sr1 and Sr2.

  The processing in steps S120 to S128 is the same as the processing in the first embodiment shown in FIG. That is, in steps S120 to S126, the interval calculator 23 acquires each time interval Δs based on the spots Sr1 to Srx, and in step S128, the interval calculator 23 calculates an average time interval Δt of the time intervals Δs.

In step S130, the prediction determination unit 24 advances the moving image M by the average time interval Δt and increments the count value m by 1.
In step S133, the prediction determination unit 24 determines whether or not a spot Sr (x + m) exists in the mask region 33 at the time tx + m × Δt. When the spot Sr (x + m) exists (step S133: YES), in step S134, the prediction determination unit 24 stores the count value m and the determination value d = 1 in the memory 26. On the other hand, when the spot Sr (x + m) does not exist (step S133: NO), in step S136, the prediction determination unit 24 stores the count value m and the determination value d = 0 in the memory 26.
The processing in step S138 is the same as the processing in the first embodiment shown in FIG. That is, when the prediction determination ends, the process proceeds to step S140.

  The processes in steps S140, S150, and S155 are the same as those in the first embodiment shown in FIG. That is, in step S140, the astigmatism specifying unit 25 determines whether or not the discrimination value d = 1 for all count values m. When the discriminant value d = 1 for all count values m (step S140: YES), the astigmatism specifying unit 25 causes the output unit 30 to output all lighting information in step S150. On the other hand, when the discriminant value d = 0 for the specific count value m = y (step S140: NO), in step S155, the astigmatism specifying unit 25 determines that the illuminating lamp 2- (x + y) is based on the count value y. The output unit 30 is caused to output astigmatism information indicating that it is a defect. As a result, the spotlight 2 is detected and specified.

  As described above, in the astigmatism detection device 1 according to the present embodiment, the area setting unit 22 sets the mask area 33 as the detection area, and the mask area 33 includes the illumination lamp 2-1 and the illumination lamp 2-2. It is defined by the image portion indicating the installation interval. As described above, the mask region 33 is defined by the position interval between the spot Sr1 and the spot Sr2 corresponding to the illumination lamp 2-1 and the illumination lamp 2-2. Accordingly, since the presence / absence of the spot Sr is determined using the mask area 33 having a relatively wide width in the moving direction of the spot Sr, even if the detection timing by the prediction determination unit 24 is slightly deviated, it becomes a detection target. The mask region 33 is surely included in the spot Sr or the position that should be the spot (that is, the position corresponding to the spotlight 2). Further, since the width of the mask region 33 does not include two or more spots Sr, even if the detection timing by the prediction determination unit 24 is slightly shifted, the detection timing order and the detection target spot Sr or spot There is no discrepancy in the order of positions that should be. Therefore, it is possible to reliably determine the incongruity for each of the illumination lamps 2.

<Programs>
In addition, each component which implement | achieves the point detection apparatus 1 (CPU21) in each embodiment mentioned above, and each step of a process are implement | achieved when the program memorize | stored in RAM or ROM of the memory 26 operate | moves. .

  Further, according to the present invention, a software program (a program corresponding to the flowchart shown in FIG. 5 or FIG. 8) that realizes the functions of the above-described embodiments is directly or remotely supplied to the defect detection device 1 (CPU 21). This includes cases where Therefore, the program code itself installed in the defect detection device 1 (CPU 21) in order to realize the functional processing of the present invention is also included in the present invention. That is, the present invention includes a computer program for realizing the functional processing of the present invention. The program can cause the computer to function as all or part of the region setting unit 22, the interval calculation unit 23, the prediction determination unit 24, and the astigmatism specifying unit 25. As described above, according to the present invention, the operational effects as shown in the above-described embodiment can be realized by the introduction of software, so that the ease of introduction of the defect detection device 1 can be improved.

  The program may be downloaded from the Internet. In this case, the computer program or the compressed file including the automatic installation function is downloaded to the hard disk or the like from the homepage to the incongruity detection apparatus 1 connected to the homepage on the Internet by the browser function.

  A computer-readable storage medium storing the above program is also included in the present invention. When the program is supplied by a storage medium, the storage medium may be a flexible disk, hard disk, optical disk, magneto-optical disk, etc. MO, CD-ROM, CD-R, CD-RW, magnetic tape, non-volatile May be a memory card, ROM, DVD (DVD-ROM, DVD-R) or the like.

<Modification>
Although the preferred embodiment of the present invention has been described above, the present invention can be modified in various ways as described below.

(1) Modified Example of First Embodiment In the first embodiment, the configuration in which astigmatism detection is predicted based on the average time interval Δt related to the spot Sr is shown, but each time related to the spot Sr. A modification in which astigmatism detection is statistically performed based on the interval Δs is also possible.

  In this modification, the prediction determination unit 24 is not used in the image analysis unit 20. The interval calculation unit 23, for all spots Sr, Δsk = difference between the time tk when the spot Srk passes through the detection region 32 and the time t (k-1) when the spot Sr (k−1) passes through the detection region 32. tk−t (k−1) is calculated and stored in the memory 26. The astigmatism specifying unit 25 determines the unit time interval Δs0 that is the minimum time interval from the stored time intervals Δs2 to Δsk. Alternatively, the interval Δs2 between the spots Sr1 and Sr2 may be set as the unit time interval Δs0. The point identifying unit 25 determines whether or not all the time intervals Δs are substantially equal to the unit time interval Δs0. This determination can be made, for example, by determining whether or not the target time interval Δs is included within a predetermined ratio (for example, 1.5 times) with respect to the unit time interval Δs0.

  When all the time intervals Δs are substantially equal to the unit time interval Δs 0, the astigmatism specifying unit 25 provides the output unit 30 with all lighting information indicating that all of the lamps 2-1 to 2-n are lit. Output. On the other hand, when the time interval Δsz of the specific count value k = z is close to a multiple of the unit time interval Δs0, the astigmatism specifying unit 25 sets the illumination lamp 2 after the illumination lamp 2- (z−1). Determine that there is a defect. For example, when the time interval Δsz of the specific count value k = z is around twice the unit time interval Δs0, the astigmatism specifying unit 25 determines that the illumination lamp 2-z is astigmatic. In addition, when the time interval Δsz of the specific count value k = z is close to three times the unit time interval Δs0, the astigmatism specifying unit 25 determines that the illumination lamps 2-z and 2- (z + 1) are astigmatism. It is determined that Then, as in the above embodiments, the astigmatism specifying unit 25 causes the output unit 30 to output astigmatism information regarding the illuminating lamp 2 that is an astigmatism.

  In this way, the astigmatism specifying unit 25 can specify that the illumination lamp 2 corresponding to the period included in the time interval Δsz is astigmatic when the time interval Δsz is a multiple of the unit time interval Δs0. . According to this modification, the prediction determination unit 24 is not required, and the configuration of the astigmatism detection device 1 is further simplified. Moreover, even when the lighting lamp 2 close to the entrance Ta has a spot, the spot detection process can be performed.

(2) Modification of Second Embodiment In the second embodiment described above, a configuration is shown in which whether or not the spot Sr is included in the mask region 33 is determined for each average time interval Δt. A configuration in which the presence of a defect is determined when there is a period in which no spot is included in the region 33 is also possible. That is, since the mask area 33 is defined by the width of the spot Sr1 and the spot Sr2, if two adjacent illumination lamps 2 are both lit, one of the two illumination lamps 2 is necessarily mask area 33. Will be included. On the other hand, when one of the two adjacent illumination lamps 2 is inconspicuous, a period in which no spots are included in the mask region 33 occurs. When there is a spot absence period, the astigmatism specifying unit 25 can specify that the corresponding illumination lamp 2 is astigmatism.

  However, it should be noted that when a plurality of spotlights are consecutive, the existence of the spotlight can be specified, but which lamp is not spotted. On the other hand, in this modification, since the astigmatism detection process is based on the positional periodicity of the illuminating lamp 2 (that is, the time element is not included in the astigmatism detection process), the traveling speed of the vehicle V is substantially constant. There is no need to. Therefore, a simpler defect detection method is realized.

(3) Modifications related to illuminating lamps 2 and 3 In each of the above embodiments, the illuminating lamp 2 above the right lane and the spot Sr astigmatism detection processing have been described. It is also applicable to the spot Su. Further, the astigmatism detection process for the illumination lamp 2 and the astigmatism detection process for the illumination lamp 3 may be executed simultaneously in parallel. Further, in each of the above embodiments, the configuration in which the illumination lamp 2 is a tunnel lamp is shown, but the illumination lamp 2 is not limited to this and may be a road lamp installed on a general road.

(4) Modifications related to moving image M In the above embodiments, the configuration in which the image analysis process (astigmatism detection process) is applied to the captured image acquired from the in-vehicle camera C has been described. The same image analysis process as described above may be applied to the luminance distribution image generated by converting the corresponding color or light and shade. In this case, in the second embodiment, the auxiliary line for setting the mask region 33 is a line connecting the centers of the spots Sr1 and Sr2, that is, two maximum luminance points that appear at the upper right end of the screen first. What is necessary is just to be defined as a line which connects.

(5) Deformation of shooting direction of moving image M In each of the above-described embodiments, the configuration in which the in-vehicle camera C captures the front of the vehicle V in the traveling direction is shown. Good. In this case, the same astigmatism detection process as described above may be performed on the reverse video of the captured video.

DESCRIPTION OF SYMBOLS 1 Astigmatism detection apparatus 2, 2-1 to 2-n, 3, 3-1 to 3-n 'Illumination lamp 10 Image acquisition part 20 Image analysis part 21 CPU
22 region setting unit 23 interval calculation unit 24 prediction determination unit 25 astigmatism specifying unit 26 memory 30 output unit 31 display screen 32 detection region 33 mask region (detection region)
35 Operation Input Unit M Movie Sr, Sr1 to Srn, Su, Su1 to Sun 'Spot V Vehicle (moving object)

Claims (14)

An astigmatism detection device for detecting astigmatism of an illumination lamp,
An image acquisition unit for acquiring a moving image including a plurality of illumination lamps photographed during traveling of the moving object;
An image analysis unit for identifying a spotlight among the plurality of illumination lights based on a periodicity of spots corresponding to the plurality of illumination lights, which passes through a detection region in a display screen for displaying the moving image; An incongruity detection device. 2. The astigmatism detection device according to claim 1, wherein the moving body travels at a substantially constant speed, and the first to n-th spots corresponding to the plurality of illumination lights pass through the detection area in this order. ,
The image analysis unit
An area setting unit for setting the detection area;
An interval calculation unit for calculating an average time interval in which the first to x-th spots pass through the detection region;
A prediction determination unit that determines whether or not spots after the (x + 1) -th appear in the detection region at every average time interval;
An astigmatism detection device comprising: an astigmatism specifying unit that specifies an unsatisfactory illumination lamp based on the fact that the x + 1th and later spots do not appear in the detection area at every average time interval.   3. The astigmatism detection device according to claim 2, wherein the detection area is set at an upper right end portion or an upper left upper end portion of the display screen, and a count value m × the time from a time when the x-th spot passes through the detection area. Disadvantage configured such that, when the x + m spot does not exist in the detection area at the time when the average time interval has elapsed, the discontinuity specifying unit specifies that the x + m illumination lamp is inconspicuous. Detection device. 2. The astigmatism detection device according to claim 1, wherein the moving body travels at a substantially constant speed, and the first to n-th spots corresponding to the plurality of illumination lights pass through the detection area in this order. ,
The image analysis unit
An area setting unit for setting the first and second detection areas as the detection area;
An interval calculation unit for calculating an average time interval in which the first to x-th spots pass through the first detection region;
A prediction determination unit that determines whether or not spots after the (x + 1) -th appear in the second detection area at every average time interval;
A non-spot identifying unit that identifies a non-spot illuminating lamp based on the fact that the spot after the x + 1th does not appear in the second detection area at every average time interval, and at least the second detection The astigmatism detection device, wherein the region is defined by an image portion indicating an installation interval between the first illumination lamp and the second illumination lamp.   5. The astigmatism detection device according to claim 4, wherein the second detection region is the second detection region at the time when the count value m × the average time interval has elapsed from the time when the x-th spot has passed through the first detection region. An astigmatism detection device configured to specify that the x + m-th illuminating lamp is astigmatic when the x + m spot does not exist. The astigmatism detection device according to claim 1, wherein the moving body travels at a substantially constant speed,
The image analysis unit
An interval calculation unit for calculating a time interval for the spot to pass through the detection region;
A point identifying unit that identifies that the lamp corresponding to the period included in the specific time interval is a point when the specific time interval calculated by the interval calculation unit is a multiple of another time interval. An incongruity detection device.   A program for causing a computer to function as each unit included in the astigmatism detection device according to any one of claims 1 to 6.   A computer-readable storage medium in which the program according to claim 7 is stored. An astigmatism detection method for detecting an astigmatism of an illumination lamp using an astigmatism detection device having an image acquisition unit and an image analysis unit,
An image acquisition step of acquiring a moving image including a plurality of illumination lamps photographed when the moving body travels by the image acquisition unit;
Based on the periodicity of spots corresponding to the plurality of illumination lamps that pass through a detection area in the display screen for displaying the moving image by the image analysis unit, the inconsistent illumination lamps of the plurality of illumination lamps An astigmatism detection method comprising an image analysis step to be identified. 10. The astigmatism detection method according to claim 9, wherein the moving body travels at a substantially constant speed, and first to n-th spots corresponding to the plurality of illumination lights pass through the detection area in this order. ,
The image analysis step includes
An area setting step for setting the detection area;
An interval calculating step for calculating an average time interval in which the first to xth spots pass through the detection region;
A prediction determination step for determining whether or not spots after the (x + 1) -th appear in the detection area at every average time interval;
An astigmatism detection method including an astigmatism specifying step of specifying an unsatisfactory illumination lamp based on the fact that the spot after the (x + 1) th does not appear in the detection area at every average time interval.   11. The astigmatism detection method according to claim 10, wherein the detection area is set at an upper right end portion or an upper left end portion of the display screen, and a count value m × the time from a time when the x-th spot passes through the detection area. An astigmatism detection method in which, when the x + m-th spot does not exist in the detection area at the time when the average time interval has elapsed, the x + m illumination lamp is identified as an astigmatism in the astigmatism specifying step. 10. The astigmatism detection method according to claim 9, wherein the moving body travels at a substantially constant speed, and first to n-th spots corresponding to the plurality of illumination lights pass through the detection area in this order. ,
The image analysis step includes
An area setting step for setting the first and second detection areas as the detection area;
An interval calculating step for calculating an average time interval in which the first to x-th spots pass through the first detection region;
A prediction determination step of determining whether or not spots after the (x + 1) -th appear in the second detection area at every average time interval;
An astigmatism identifying step of identifying an unsatisfactory illumination lamp based on the fact that the spot after the x + 1th does not appear in the second detection area at every average time interval, and at least the second detection The astigmatism detection method, wherein the region is defined by an image portion indicating an installation interval between the first illumination lamp and the second illumination lamp.   13. The astigmatism detection method according to claim 12, wherein the second detection region is the second detection region at the time when the count value m × the average time interval has elapsed from the time when the x-th spot has passed through the first detection region. An astigmatism detection method in which, when there is no x + m spot, the x + m-th illuminating lamp is identified as an astigmatism in the astigmatism specifying step. The astigmatism detection method according to claim 9, wherein the moving body is driven at a substantially constant speed,
The image analysis step includes
An interval calculation step for calculating a time interval for the spot to pass through the detection area;
An astigmatism identifying unit that identifies that the illumination lamp corresponding to the period included in the specific time interval is astigmatic when the specific time interval calculated by the interval calculation step is a multiple of another time interval. And a method for detecting an incongruity including a step.

Images