Contact wire testing device and contact wire testing method

Abstract

A contact wire detection device and method are provided, which accurately detect the wear area of ​​the contact wire. A linear sensor image generation unit 30 generates a linear sensor image from an image captured by a linear sensor 22. A wear area image generation unit 32 performs image processing according to a plurality of image processing execution programs M registered in an image processing execution program memory unit 33 to capture the wear area of ​​the contact wire. It then corrects the wear area by scanning outwards from the captured wear area, using a position where the brightness value is below a preset brightness value as the boundary of the wear area, generating a plurality of wear area captured images corresponding to each of the plurality of image processing execution programs M. A wear area captured image evaluation unit 34 selects the wear area captured image with the highest evaluation value from the plurality of wear area captured images. A contact wire offset calculation unit 35 and a contact wire residual diameter calculation unit 36 ​​calculate the contact wire offset and residual diameter based on the selected wear area captured image.

Description

Contact line detection device and contact line detection method The present invention relates to a contact line detection device and a contact line detection method. When an electric railway vehicle is traveling on the line, it obtains power by supplying electricity from the contact wire installed above the line through a pantograph or other current collecting devices. Every time an electric railway vehicle passes, the current collecting device slides relative to the bottom of the contact wire. Therefore, if the electric railway vehicle continues to operate, the contact wire gradually wears out and eventually breaks. Therefore, a wear limit is set for the contact wire, and a new contact wire is replaced when the residual diameter of the contact wire is lower than the management value. On the other hand, since the contact part of the current collecting device also wears out, in order to prevent the contact part from being concentrated in one place, the contact wire is arranged in a zigzag deviation along the right angle of the track, and the deviation amount is also managed. When the operator manually performs the inspection and measurement of the contact wire, it takes a lot of man-hours. For example, Patent Document 1 discloses an automatic detection device comprising: lighting that illuminates the wear area located below the contact wire above an electric railway vehicle; and a linear sensor that takes an image of the wear area of ​​the contact wire (the sliding surface in contact with the pantograph) illuminated by the lighting. More specifically, the linear sensor is set up so as to face upward and have a scanning line that crosses the contact wire, and the images taken by the linear sensor are arranged in a time sequence to generate a linear sensor image. Then, after image processing is applied to the linear sensor image, the deviation, residual diameter, etc. of the contact wire are calculated by using known data related to the specifications of the contact wire, such as the height and thickness of the contact wire. [Prior Art Document] [Patent Document] Patent Document 1: Japanese Patent No. 5287177 (Problem to be Solved by the Invention) However, in the above-mentioned device, due to factors such as lighting, the state of the wear area of ​​the contact line, and the photographic environment, there may be differences in the brightness of the wear area of ​​the contact line in the linear sensor image, resulting in the problem of not being able to correctly obtain the wear area. Therefore, the present invention aims to provide a contact wire detection device and a contact wire detection method that can accurately detect the wear area of ​​the contact wire. (Technical Means for Solving the Problem) In order to solve the above problems, the present invention adopts the following means. That is, the contact line detection device of the first invention is provided with a linear sensor for photographing the contact line, and uses the image photographed by the linear sensor as an input image to detect the above contact line; it is characterized in that it has: a linear sensor image generating unit, which arranges the above input image in time sequence to generate a linear sensor image; an image processing execution program memory unit, which logs a plurality of execution orders as image processing execution programs when executing a plurality of image processing functions in sequence on the linear sensor image; a wear area capture image generating unit, which uses the above linear sensor image as input to execute the above plurality of image processing execution programs according to each of the above plurality of image processing execution programs. An image processing function is used to capture the wear area of ​​the above-mentioned contact line, and by scanning outward from the wear area of ​​the above-mentioned captured contact line, the position where the brightness value becomes below a preset brightness value is used as the boundary of the wear area, and the above-mentioned wear area is corrected to generate a plurality of wear area capture images corresponding to each of the above-mentioned plurality of image processing execution programs; a wear area capture image evaluation unit, which calculates the evaluation value of each of the above-mentioned plurality of wear area capture images and selects the wear area capture image with the highest evaluation value; and a contact line detection unit, which calculates either or both of the residual diameter and the deviation of the above-mentioned contact line based on the above-mentioned selected wear area capture image with the highest evaluation value. In addition, the contact line detection device of the second invention is characterized in that for each of the above-mentioned multiple image processing execution programs, contrast adjustment processing of 0 or more, background / noise removal processing of 0 or more, binarization processing of 1 or more, and interpolation processing of 0 or more are logged in a sequential execution manner. In addition, the contact line detection device of the third invention is characterized in that a priority is established and memorized correspondingly for each of the above-mentioned multiple image processing execution programs, and when the same evaluation value is calculated for multiple above-mentioned wear area capture images, the above-mentioned wear area capture image evaluation unit selects the above-mentioned wear area capture image corresponding to the above-mentioned image processing execution program with a higher priority as the above-mentioned wear area capture image with the highest evaluation value. In addition, the contact line detection device of the fourth invention is characterized in that when the number of the above-mentioned wear areas is calculated for each of the above-mentioned multiple wear area capture images, and when the difference between the number of the above-mentioned wear areas and the actual number of the above-mentioned contact lines is small, the above-mentioned wear area capture image evaluation unit calculates the above-mentioned evaluation value in a manner that the above-mentioned evaluation value becomes higher. In addition, the contact line detection device of the fifth invention is characterized in that the above-mentioned wear area captured image evaluation unit calculates the contact line likelihood for each of the above-mentioned multiple wear area captured images, and when the contact line likelihood is higher, the above-mentioned evaluation value is calculated in a manner that the above-mentioned evaluation value becomes higher, wherein the contact line likelihood is an indicator showing that the above-mentioned wear area has a feature that is similar to the above-mentioned contact line on the image. In addition, the contact line detection device of the 6th invention is characterized in that the above-mentioned image processing execution program memory unit has the above-mentioned image processing execution program logged therein corresponding to each of the multiple different environments when the above-mentioned image is captured by the above-mentioned linear sensor, and further has an environment judgment unit, which judges the above-mentioned environment in which the above-mentioned input image is captured based on the above-mentioned linear sensor image, and the above-mentioned wear area capture image generation unit generates the above-mentioned wear area capture image in accordance with the above-mentioned image processing execution program corresponding to the judged above-mentioned environment. In addition, the contact line detection method of the 7th invention is a contact line detection method that uses an image captured by a linear sensor that captures the contact line as an input image and detects the above-mentioned contact line based on this image; it is characterized in that it includes the following steps: arranging the above-mentioned input image in time sequence to generate a linear sensor image; when executing a plurality of image processing functions on the linear sensor image in sequence, logging a plurality of their execution orders as image processing execution programs, and executing the above-mentioned plurality of image processing functions using the above-mentioned linear sensor image as input according to each of the plurality of image processing execution programs, thereby The wear area of ​​the contact line is captured; scanning outward from the captured wear area of ​​the contact line, using the location where the brightness value falls below a predetermined brightness value as the boundary of the wear area to correct the wear area, thereby generating a plurality of wear area capture images corresponding to each of the plurality of image processing execution programs; calculating an evaluation value for each of the plurality of wear area capture images, selecting the wear area capture image with the highest evaluation value; and calculating one or both of the residual diameter and the deviation of the contact line based on the selected wear area capture image with the highest evaluation value. (Compared to the effectiveness of the prior art) According to the present invention, even when a brightness difference occurs in the worn area, the worn area of ​​the contact line can be accurately detected. The following describes an embodiment of the present invention with reference to the accompanying drawings. FIG1 is an illustrative diagram showing the driving environment of an inspection vehicle 1 equipped with a contact wire detection device 20 according to this embodiment and performing contact wire inspection. In FIG1 , the inspection vehicle 1 travels on a pair of ground-mounted lines 2. A plurality of poles 3 are erected at intervals along the lines 2 on either side of the lines 2. A crossbeam 5 and a curved traction metal member 6 are provided on the poles 3. A power line 4, or feeder line 4, supplying power to the inspection vehicle 1 is provided above the crossbeam 5, extending in the direction R of the lines 2 and supported by the crossbeam 5. Below the crossbeam 5 and the curved traction metal member 6, a suspension wire 7, from which a contact wire 9 (described below) is suspended, is supported by the crossbeam 5 and the curved traction metal member 6, similarly extending in the direction R of the lines 2. A plurality of hangers 8 are suspended from the suspension wire 7 at approximately regular intervals along the direction R of the lines 2. Contact wires 9 are suspended from a plurality of hangers 8 with a certain tension and are located below the suspension wires 7, generally parallel to the suspension wires 7. Power is supplied to the contact wires 9 from the feeder line 4 via a feeder branch line (not shown). A power collector 1b, such as a pantograph, is installed on the top surface 1a of the inspection vehicle 1. While the inspection vehicle 1 is in motion, the power collector 1b contacts the contact wires 9, thereby supplying power from the contact wires 9 to the inspection vehicle 1. One or two contact wires 9 are arranged parallel to a pair of tracks 2. During operation of an electric railway vehicle, such as the inspection vehicle 1, when a contact wire 9, serving as a power supply source, moves toward the next adjacent contact wire 9 in the extension direction R of the tracks 2, to prevent a section where power is not supplied, the contact wires 9 are arranged so that their ends overlap with other adjacent contact wires 9 along the extension direction R of the tracks 2. Therefore, in this embodiment, there may be sections where a maximum of four contact wires 9 are arranged overlapping with each other in the pair of tracks 2. Furthermore, when imaging a branching portion of the contact wires 9 corresponding to a track branch, it is possible to image a greater number of contact wires 9 than in a typical section. If contact wire 9 is installed parallel to the lines 2, the portion of the current collector 1b that contacts the contact wire 9 is limited. This creates a risk of the current collector 1b shifting, causing wear and damage. To prevent this, contact wire 9 is installed in a gently zigzagging pattern along the extension direction R of the lines 2. The left-right deviation of contact wire 9 from the center between the pair of lines 2, or misalignment, caused by this zigzagging is strictly controlled. Every time an electric railway vehicle, such as the inspection vehicle 1, passes, the current collector 1b slides against the underside of the contact wire 9. Consequently, if the electric railway vehicle continues to operate, the contact wire 9 gradually wears and eventually breaks. Therefore, a wear limit is set for the contact wire 9. When the remaining diameter of the contact wire 9 falls below the control value, the contact wire 9 is replaced with a new one. The contact wire detection device 20 of this embodiment detects the contact wire 9 by measuring its residual diameter or displacement. The contact wire detection device 20 includes an illumination device 21, a linear sensor 22, and a control device 23. The illumination device 21 is positioned on the upper surface 1a of the inspection vehicle 1 so as to illuminate upward. When the linear sensor 22 (described below) captures an image of the contact wire 9, the image captures the worn area on the lower side of the contact wire 9, which has been formed into a flat surface due to the sliding of the current collector 1b, as reflected light from the illumination device 21. Similar to the lighting 21, the linear sensor 22 is installed on the upper surface 1a of the inspection vehicle 1. The linear sensor 22 is a camera that captures an image consisting of one pixel in the vertical direction and a plurality of pixels in the horizontal direction, for example, 8192 pixels. In this embodiment, since the linear sensor 22 captures grayscale images, each pixel has a single brightness value. The linear sensor 22 is arranged so as to face vertically upward, with the direction 22b of its scanning line 22a perpendicular to the extension direction R of the track 2 and aligned with the extension direction of the sleepers (not shown). Consequently, the scanning line 22a of the linear sensor 22 is configured to traverse the contact wire 9. The linear sensor 22 continuously captures images of the contact wire 9 at short intervals while the inspection vehicle 1 is traveling. The images captured by the linear sensor 22 are transmitted as input images to the control device 23. The control device 23 is an information processing device such as a personal computer. The control device 23 is mounted on the inspection vehicle 1. FIG2 is a block diagram illustrating the functional configuration of the contact line detection device 20 according to this embodiment. In FIG2 , the control device 23 includes a linear sensor image generator 30, an environment determination unit 31, a worn area captured image generator 32, an image processing execution program storage unit 33, a worn area captured image evaluation unit 34, a contact line deviation calculator (contact line detection unit) 35, and a contact line residual diameter calculator (contact line detection unit) 36. The linear sensor image generating unit 30 receives an input image from the linear sensor 22 and arranges the input image into, for example, 1000 lines in a time-sequential manner in the vertical direction to generate a linear sensor image. FIG3 is a schematic diagram showing an example of a linear sensor image generated from an image captured by the linear sensor of the contact line detection device according to this embodiment. In the linear sensor image 50, the horizontal direction X represents the direction 22b of the scan line 22a of the linear sensor 22, and the vertical direction Y represents the time sequence direction, i.e., the direction in which the input image is arranged. In the linear sensor image 50 shown in Figure 3, the contact wire 9 is captured at the foreground. In its new state, the contact wire 9 has a roughly circular cross-section. Its underside is worn away by the sliding motion of the current collector 1b, forming a planar wear region 9a. Because the wear region 9a is illuminated by the illumination 21, it is captured brightly. On both sides of the wear region 9a, the surface 9b of the contact wire 9 is captured, extending upward from the widthwise ends of the wear region 9a. In the linear sensor image 50 of Figure 3, the portion of the contact wire 9 bent by the bending of the pull metal member 6 is captured. Furthermore, due to factors such as the illumination, the state of the wear region of the contact wire, and the photographic environment, brightness differences may occur within the wear region 9a, resulting in dark, locally low-brightness areas 9c and 9d. The image shows the suspension wire 7 deep behind the contact wire 9. Further behind the suspension wire 7, the feeder wire 4, the beam 5 supporting it, and the curved traction metal fitting 6 are captured. Due to the imaging interval of the linear sensor 22, the beam 5 and the curved traction metal fitting 6 appear thinner than they actually are. A blank space 10 is captured against the backdrop of the contact wire 9, suspension wire 7, feeder wire 4, beam 5, and curved traction metal fitting 6. In this linear sensor image 50, because it was captured during bright daylight hours, the blank space 10 appears brightly with a high brightness value. The linear sensor image generating unit 30 generates the linear sensor image 50 as described above and transmits the image to the worn area captured image generating unit 32 . The environment determination unit 31 determines the environment in which the input image was captured based on the linear sensor image 50 generated by the linear sensor image generation unit 30. The environment determination unit 31 determines five types of environments: night or tunnel, dim daytime, bright daytime, the moment of entering a tunnel during the daytime, and other conditions. To make this determination, the environment determination unit 31 calculates a histogram of the linear sensor image 50 and calculates the average, the highest frequency value, the second-highest frequency brightness value, and the third-highest frequency brightness value. The environment is then determined by comparing the calculated values ​​with the pre-set nighttime determination threshold, dim daytime determination threshold, and bright daytime determination threshold, as described below. When all of the following judgment formulas are satisfied simultaneously, the environment judgment unit 31 judges that the environment corresponding to the linear sensor image 50 is night or a tunnel. 0≦Average value≦Night judgment threshold 0≦Highest frequency value≦Night judgment threshold 0≦Second highest frequency brightness value≦Night judgment threshold 0≦Third highest frequency brightness value≦Night judgment threshold When all of the following judgment formulas are satisfied simultaneously, the environment judgment unit 31 judges that the environment corresponding to the linear sensor image 50 is a dark daytime environment. Night judgment threshold < average value ≦ dark daytime judgment threshold Night judgment threshold < highest frequency value ≦ dark daytime judgment threshold Night judgment threshold < 2nd highest frequency brightness value ≦ dark daytime judgment threshold Night judgment threshold < 3rd highest frequency brightness value ≦ dark daytime judgment threshold When all of the following judgment formulas are satisfied simultaneously, the environment judgment unit 31 judges that the environment corresponding to the linear sensor image 50 is bright daytime. Bright daytime judgment threshold value < average value Bright daytime judgment threshold value < highest frequency value Bright daytime judgment threshold value < second highest frequency brightness value Bright daytime judgment threshold value < third highest frequency brightness value When all of the following judgment formulas are satisfied simultaneously, the environment judgment unit 31 judges that the environment corresponding to the linear sensor image 50 is the moment of entering the tunnel during the day. Night judgment threshold < average value ≦ maximum brightness value (e.g., 255) 0 ≦ highest frequency value ≦ night judgment threshold 0 ≦ second highest frequency brightness value ≦ night judgment threshold 0 ≦ third highest frequency brightness value ≦ night judgment threshold If none of the above environments are met, the environment determination unit 31 determines that the environment is another environment and transmits the determined environment to the wear area captured image generation unit 32 . The wear area capture image generation unit 32 receives the linear sensor image 50 from the linear sensor image generation unit 30 and the determination result (environment) from the environment determination unit 31, and performs various image processing functions on the linear sensor image 50 to capture the wear area 9a of the contact line 9 and generate a wear area capture image. The worn area captured image generation unit 32 includes an image processing unit 40 and a post-processing unit 41. The image processing unit 40 is configured to selectively perform various image processing functions. The image processing unit 40 includes a contrast adjustment unit 40A, a background / noise removal unit 40B, a binarization unit 40C, an interpolation unit 40D, and a worn area correction unit 40E. The contrast adjustment processing unit 40A is configured to selectively execute one or more image processing functions for adjusting image gradation. Examples of contrast adjustment processing include gamma correction, which adjusts image gradation to an appropriate curve corresponding to the gamma value, or gradation adjustment by normalizing brightness values. Furthermore, the contrast adjustment processing included in the contrast adjustment processing unit 40A is not limited to the aforementioned processing. The background / noise removal processing unit 40B is configured to selectively execute one or more image processing functions designed to remove background objects or noise, such as buildings, that differ from overhead lines. A suitable background / noise removal process is GSTH (grayscale capping), which involves performing an opening operation (e.g., shrinking or dilating a grayscale image) and then subtracting the source image from the expansion / contraction process. The background / noise removal processing unit 40B includes more than just the aforementioned processing. The binarization unit 40C is configured to selectively execute one or more image processing functions. These image processing functions are intended to convert a multi-layered image into two layers, for example, white and black, and to extract only the worn area 9a. For example, discriminant analysis binarization can be applied as the binarization process. This discriminant analysis binarization process generates a histogram of brightness values ​​for the image and determines a binarization threshold based on the image to maximize the ratio of the within-class variance to the between-class variance for each of the background and the worn area 9a. The binarization processes included in the binarization unit 40C are not limited to those described above. The interpolation processing unit 40D is configured to selectively execute one or more image processing functions. The image processing function is to interpolate the shape of the worn area 9a by connecting the areas after the binarization processing in the binarization processing unit 40C is executed. The image of the area assumed to be the worn area 9a is captured as the object, and the area originally assumed to be one worn area 9a is divided into a plurality of areas. Applicable interpolation processes include closing processing and wave wear interpolation processing. Closing processing connects areas that are displayed in a similar manner to the horizontal direction X by expanding or contracting the image in the horizontal direction X. Wave wear interpolation processing connects areas that are displayed in a striped or wavy pattern in the vertical direction Y. The interpolation processing included in the interpolation processing unit 40D is not limited to the processing described above. FIG4 is a schematic diagram showing an example of a temporarily captured wear area image obtained by the contact line detection device of this embodiment. This is a captured wear area image 50A processed by the interpolation processing unit 40D as described above. In the captured wear area image 50A, due to factors such as lighting, the state of the worn area 9a of the contact line 9, and the photographic environment, a brightness difference may occur in the worn area 9a, resulting in dark areas 9c and 9d with lower brightness than other areas, as shown in FIG3 . In other words, in the captured wear area image 50A processed by the interpolation processing unit 40D, the dark areas 9c and 9d are not detected as the worn area 9a, and therefore the correct worn area 9a cannot be detected. As shown in Figure 4, in the worn area captured image 50A processed by the interpolation processing unit 40D, only the brightly displayed worn area 9a is captured. Therefore, in this embodiment, the worn area captured image 50A processed by the interpolation processing unit 40D is corrected so that dark areas 9c and 9d are detected as worn area 9a. The worn area correction unit 40E uses raster scanning to search outward from the position of the worn area captured for each vertical line in the image during the wear area 9a capture process. The position where the brightness value falls below a predetermined brightness value is defined as the boundary between the sliding surface and the non-sliding surface, that is, the boundary of the worn area 9a. FIG5 is a schematic diagram showing an example of a wear area capture image ultimately obtained by the contact line detection device of this embodiment. As a result of the correction described above, the dark areas 9c and 9d shown in FIG4 are detected as wear areas 9a. Therefore, as shown in FIG5 , the wear areas 9a are correctly detected. The wear area correction unit 40E transmits the corrected wear area capture image 50B to the post-processing unit 41, along with the number of corresponding wear areas 9a, information about the length of each wear area 9a in the longitudinal direction Y, including the start and end coordinates, and information regarding the image processing execution program (or processing mode) M corresponding to the wear area capture image. In this way, the wear area capture image generation unit 32, through the image processing unit 40 and the post-processing unit 41, and in accordance with each of the plurality of image processing execution programs M, takes the linear sensor image 50 as input and performs image processing functions, thereby correctly capturing the wear area of ​​the contact line 9 and generating a plurality of wear area capture images 50B corresponding to each of the plurality of image processing execution programs M. The image processing functions included in the contrast adjustment processing unit 40A, background / noise removal processing unit 40B, binarization processing unit 40C, interpolation processing unit 40D and wear area correction unit 40E are executed in accordance with the execution program when executing each image processing function, that is, the processing mode (image processing execution program) M logged in the image processing execution program memory unit 33. Fig. 6 is an explanatory diagram of the image processing execution program storage unit of the contact line detection device of this embodiment. In Fig. 6, the image processing execution program storage unit 33 includes a processing storage unit 33a, a parameter set storage unit 33b, and a processing mode registration unit 33c. The processing storage unit 33a stores various image processing functions. These image processing functions are executed in the contrast adjustment processing unit 40A, background / noise removal processing unit 40B, binarization processing unit 40C, interpolation processing unit 40D, and worn area correction unit 40E described above. The processing storage unit 33a in FIG6 stores X image processing functions PR1 to PRX included in any of the aforementioned processing units 40A, 40B, 40C, 40D, and 40E. Furthermore, the parameter set storage unit 33b stores parameter sets that can be used when actually executing each of the image processing functions PR1 to PRX stored in the processing storage unit 33a. In FIG6 , the parameter set storage unit 33b stores Y types of parameter sets, from PS1 to PSY, which are used in any one or more of the X image processing functions PR1 to PRX stored in the processing storage unit 33a. The processing mode register 33c stores a plurality of processing modes M, which are processing modes M for executing the aforementioned image processing functions when processing the linear sensor image 50. In the processing mode register 33c of FIG6 , L processing modes M1 to ML are stored. FIG6 shows the first processing mode M1 among the L processing modes M1 to ML. Processing mode M1 registers Z image processing functions P1 to PZ, from image processing function P1 to image processing function PZ, for sequential execution. Image processing function P1 in processing mode M1 is configured to reference image processing function PR1 in the processing storage unit 33a and parameter set PS2 in the parameter set storage unit 33b. This indicates that image processing function P1 in processing mode M1 actually executes image processing function PR1 stored in the processing storage unit 33a using parameter set PS2. Furthermore, in the second image processing function P2 executed in the first processing mode M1, only the image processing function PRX stored in the processing storage unit 33a is referenced, and the parameter set in the parameter set storage unit 33b is not referenced. Thus, in processing mode M, when image processing functions PR1 to PRX stored in the processing storage unit 33a are registered, it is not necessary to establish a corresponding reference to the parameter set in the parameter set storage unit 33b. Although detailed description is omitted, the processing modes M2 to ML are similar to the processing mode M1 in that they are composed of image processing functions executed by the expression reference processing storage unit 33a and the parameter set storage unit 33b. The number of image processing functions executed in each of the processing modes M1 to ML stored in the processing mode registration unit 33c is not fixed but variable. In other words, the number of image processing functions executed in each of the processing modes M1 to ML is basically different from each other. Thus, processing mode M is formed by registering a plurality of image processing functions, or a combination of image processing functions and parameter sets used in these image processing functions, through a program creation process. Thus, processing mode M, which can also be referred to as image processing execution program M, registers the execution procedures for each image processing function when processing the linear sensor image 50, along with the parameter sets that indicate how each image processing function is actually executed. Hereinafter, processing mode M will be referred to as image processing execution program M. As described above, each image processing function is categorized as processing for the purpose of performing one of contrast adjustment, background / noise removal, binarization, and interpolation. Within each image processing execution program M, the image processing functions are registered so that they are executed in the order of their respective classifications. For example, performing contrast adjustment after binarization is essentially meaningless. Therefore, after executing the image processing function included in the binarization processing unit 40C, the image processing execution program M that executes the image processing function included in the contrast adjustment processing unit 40A is generally not registered. Furthermore, in order to capture the worn area 9a, binarization processing must be performed at a minimum. In other words, the order of image processing functions is registered in the image processing execution program M. For example, the image processing function included in the binarization processing unit 40C is executed at least once, while the image processing functions included in the other processing units 40A, 40B, 40D, and 40E are executed as needed, that is, at least 0 times. In this way, in each of the plurality of image processing execution programs M, at least 0 contrast adjustment processes, at least 0 background / noise removal processes, at least 1 binarization process, and at least 0 interpolation processes are registered for sequential execution. Furthermore, the image processing functions included in each of the processing units 40A, 40B, 40C, 40D, and 40E may be registered in a sequence such that the image processing functions can be executed multiple times. For example, in the interpolation processing unit 40D, the image processing execution program M may be constructed such that, after connecting the areas divided in the horizontal direction X by the closing process, the areas divided in the vertical direction Y by the wave wear interpolation process are connected. Alternatively, the same image processing function may be logged into the image processing execution program M in a manner of executing multiple times with the same parameter set, or continuously with a changed parameter set, or interspersed with the execution of other image processing functions included in the same processing classification. A plurality of image processing execution programs M are registered to implement all combinations of processing programs between the image processing functions within the processing units 40A, 40B, 40C, 40D, and 40E. Each of these plurality of image processing execution programs M has different accuracy in capturing the worn area 9a and adaptability to background conditions such as the linear sensor image 50. Therefore, each of the plurality of image processing execution programs M is associated with a priority and stored. For example, an image processing execution program M that consistently achieves a certain level of capture accuracy for the worn area 9a without being significantly affected by background brightness is assigned a higher priority. Thus, the priority is set so that the higher the capture accuracy of the worn area 9a, the higher the priority. As described above, the image processing unit 40 executes the image processing execution programs M respectively, generates processed images corresponding to the respective image processing execution programs M, and transmits the processed images to the post-processing unit 41 . The post-processing unit 41 receives the processed image from the image processing unit 40 and performs edge extraction to capture the worn areas 9a. If multiple worn areas 9a are captured in the processed image, the length of each worn area 9a in the longitudinal direction Y is calculated. As described above, in this embodiment, up to four contact lines 9 can be overlapped. That is, a maximum of four contact lines 9 can be captured in a single linear sensor image 50. Therefore, if five or more worn areas 9a are captured, the post-processing unit 41 treats the four areas as worn areas 9a in descending order of length, and treats the remaining areas with shorter lengths as non-worn areas 9a. The post-processing unit 41 uses the edge-captured image as the wear area capture image 50A, and transmits it to the wear area capture image evaluation unit 34 together with the number of corresponding wear areas 9a, the length information of the longitudinal direction Y of each wear area 9a including the start point coordinates and the end point coordinates, and information about the image processing execution program M corresponding to the wear area capture image. The worn area captured image evaluation unit 34 receives the corrected worn area captured images 50B corresponding to each of the plurality of image processing execution programs M and associated information, calculates an evaluation value for each of the received plurality of worn area captured images 50B, and selects the worn area captured image with the highest evaluation value. The worn area captured image generation unit 32 associates a determination of the number of worn areas 9a captured in the worn area captured image with each worn area captured image and transmits the result to the worn area captured image evaluation unit 34. The worn area captured image evaluation unit 34 compares the number of worn areas 9a in each worn area captured image with the number of contact lines 9 at the actual location where the worn area captured image was captured, and calculates the evaluation value for the worn area captured image. The evaluation value for the captured wear area image utilizes the fact that the number of contact lines on the image can be obtained from pre-registered device information. A cost is calculated for each time series (line) of the image. This is done by using either a cost evaluation (lower costs, higher evaluations) or a likelihood evaluation (higher likelihoods, higher evaluations) for each of the multiple captured wear area images. The cost evaluation is performed using the captured wear area image in the lowest-cost processing mode. The likelihood evaluation is an indicator indicating that the worn area 9a has a characteristic similar to the contact line 9 on the image. For details on cost evaluation and likelihood evaluation, see Japanese Patent No. 7196806. The wear area captured image evaluation unit 34 captures the image of the wear area with the highest evaluation value based on cost evaluation or likelihood evaluation, calculates the center of gravity coordinates and wear surface width of each wear area 9a, and transmits it together with the captured image of the wear area with the highest evaluation value to the contact line deviation calculation unit 35 and the contact line residual diameter calculation unit 36. The contact line offset calculation unit 35 receives the wear area capture image selected as the highest evaluation value and calculates the offset of the contact line 9 based on it. Figure 7(a) is an illustration of the offset calculation principle. The offset D (unit: mm) is calculated using the coordinates d (unit: pixel) of the center of gravity of the wear area 9a received from the wear area capture image evaluation unit 34, and pre-set parameters: the distance hs (mm) between the line 2 and the sensing surface 22c of the linear sensor 22, the focal length f (mm) of the lens of the linear sensor 22, the length s (mm) of the sensing surface 22c of the linear sensor 22, the number p (pixels) of the linear sensor 22, and the distance h (mm) between the line 2 and the contact line 9 calculated by a measuring device other than the contact line detection device 20. It is expressed as follows. [Formula 1] D= As described above, the contact line deviation calculation unit 35 calculates the deviation of the contact line 9 and determines whether the contact line 9 is within the allowable deviation range at the actual location corresponding to the captured image of the wear area. If the contact line 9 is not within the allowable deviation range, it is determined to be an abnormality and recorded. The contact line residual diameter calculation unit 36 ​​receives the wear area capture image selected as the highest evaluation value and calculates the residual diameter of the contact line 9 based on it. Figure 7(b) is an explanatory diagram of the calculation principle of the residual diameter. In this embodiment, the residual diameter of the contact line 9, that is, the height of the contact line 9 after wear, is calculated as the residual diameter. The contact line residual diameter calculation unit 36 ​​uses the same principle as the contact line deviation calculation unit 35 and calculates the actual wear surface width W (mm) of the contact line 9 based on the center of gravity coordinates d (pixels) and the wear surface width w (pixels) of the wear area 9a received from the wear area capture image evaluation unit 34. The residual diameter Dr (mm) of the contact line 9 is expressed by the following formula using the wear surface width W (mm) and the radius of the contact line 9. [Formula 2] Dr=r+ As described above, the contact wire residual diameter calculation unit 36 ​​calculates the residual diameter of the contact wire 9 and determines whether the residual diameter is lower than the management value. If the residual diameter is lower than the management value, it is determined to be abnormal and the content is recorded. Next, FIG8 is a flowchart illustrating the operation of the contact line detection method described above, performed by the contact line detection device 20 of this embodiment. When processing begins, the linear sensor image generation unit 30 sequentially arranges the input image from the linear sensor 22 into, for example, 1000 lines in the vertical direction to generate a linear sensor image 50 as shown in FIG3 . This image is then transmitted to the wear area capture image generation unit 32 . Furthermore, the environment determination unit 31 determines the environment in which the input image was captured based on the linear sensor image 50 generated by the linear sensor image generation unit 30 (step S10 ). Next, in the worn area captured image generation unit 32, the contrast adjustment processing unit 40A adjusts the layers of the linear sensor image 50 (step S12), the background / noise removal processing unit 40B removes objects and noise that are different from the overhead line, such as buildings captured in the background (step S14), and the binarization processing unit 40C converts the image of multiple layers into two layers, such as white and black, to capture only the worn area 9a (step S16). Next, when the image of the area assumed to be the worn area 9a is captured after the binarization processing is performed in the binarization processing unit 40C, and the area originally assumed to be one worn area 9a is divided into a plurality of areas, the interpolation processing unit 40D interpolates the shape of the worn area 9a by connecting these areas (step S18). Next, the wear area correction unit 40E searches outward for the boundary between the sliding surface and the non-sliding surface from the position captured for each longitudinal line of the image in the capture process of the wear area 9a by raster scanning, and corrects the wear area capture image 50A by setting the position where the brightness value becomes lower than the preset brightness value as the sliding surface boundary to detect the correct wear area 9a (step S20). Next, the post-processing unit 41 generates the number of corresponding wear areas 9a and the length information of each wear area 9a in the longitudinal direction Y, including the starting point coordinates and the end point coordinates, as a corrected wear area capture image 50B, and transmits it together with the information about the image processing execution program M corresponding to the wear area capture image to the wear area capture image evaluation unit 34 (step S22). Next, the worn area captured image evaluation unit 34 uses the edge-captured image from the worn area captured image generation unit 32 as a worn area captured image. Based on the number of corresponding worn areas 9a, the length information of each worn area 9a in the longitudinal direction Y, including the start and end coordinates, and information about the image processing execution program M corresponding to the worn area captured image, the unit calculates an evaluation value (likelihood evaluation or cost evaluation) for each of the received plurality of worn area captured images (step S24). The generation and evaluation of the worn area captured images are executed for all image processing execution programs M stored in the image processing execution program storage unit 33. Next, for all image processing execution programs M, determine whether the generation and evaluation of the wear area capture image has been executed (step S26). If there is an unexecuted image processing execution program M (NO in step S26), transfer to step S10 to continue processing the unexecuted image processing execution program M. On the other hand, if there is no unexecuted image processing execution program M (YES in step S26), the worn area captured image evaluation unit 34 determines the image with the highest evaluation value among all the worn area captured images as the image most suitable for capturing the worn area 9a (step S28). The worn area captured image evaluation unit 34 calculates the center of gravity coordinates and the worn surface width of each worn area 9a for the worn area captured image with the highest evaluation value, and transmits this image, along with the worn area captured image with the highest evaluation value, to the contact line deviation calculation unit 35 and the contact line residual diameter calculation unit 36. Next, the contact line offset calculation unit 35 receives the captured image of the worn region selected as the highest evaluation value and calculates the offset of the contact line 9 based on it (step S30). Furthermore, the contact line residual diameter calculation unit 36 ​​receives the captured image of the worn region selected as the highest evaluation value and calculates the residual diameter of the contact line 9 based on it (step S32). This process then terminates. According to the above embodiment, a contact line detection device 20 is provided with a linear sensor 22 for photographing the contact line 9, and uses the image photographed by the linear sensor 22 as an input image to detect the contact line 9; it comprises: a linear sensor image generating unit 30, which arranges the input image in time sequence to generate a linear sensor image 50; an image processing execution program memory unit 33, which logs a plurality of execution orders as image processing execution programs M when executing a plurality of image processing functions on the linear sensor image 50 in sequence; a wear area capture image generating unit 32, which uses the linear sensor image 50 as input to execute according to each of the plurality of image processing execution programs M The image processing function is used to capture the wear area 9a of the contact line 9, and by scanning outward from the captured contact line wear area, the position where the brightness value becomes below a preset brightness value is used as the boundary of the wear area, and the above-mentioned wear area is corrected to generate a plurality of wear area capture images corresponding to each of the plurality of image processing execution programs M; a wear area capture image evaluation unit 34, which calculates the evaluation value of each of the plurality of wear area capture images and selects the wear area capture image with the highest evaluation value; a contact line deviation calculation unit 35, which calculates the deviation of the contact line 9 based on the selected wear area capture image; and a contact line residual diameter calculation unit 36, which calculates the residual diameter of the contact line 9. In addition, the contact line detection method of the present embodiment is a method for detecting the contact line 9 by taking an image captured by the linear sensor 22 that captures the contact line 9 as an input image, and the method includes the following steps: arranging the input image in a time sequence to generate a linear sensor image 50, and when executing a plurality of image processing functions on the linear sensor image 50 in sequence, registering a plurality of execution orders as image processing execution programs M, and executing image processing in accordance with each of the plurality of image processing execution programs M using the linear sensor image 50 as input. The wear area 9a of the contact line 9 is captured by a processing function, and the captured wear area 9a is scanned outward, and the position where the brightness value becomes below the preset brightness value is used as the boundary of the wear area 9a, so as to correct the wear area 9a, generate a plurality of wear area capture images corresponding to each of the plurality of image processing execution programs M, calculate the evaluation value of each of the plurality of wear area capture images, and select the wear area capture image with the highest evaluation value, and calculate both the residual diameter and the deviation of the contact line 9 based on the selected wear area capture image. According to the above-mentioned structure, even when there is a brightness difference in the wear area 9a due to factors such as lighting, the state of the wear area 9a of the contact line 9, and the shooting environment, the wear area 9a of the contact line 9 can be correctly detected, which can improve the capture accuracy of the wear area 9a of the contact line 9 and the calculation accuracy of the residual diameter and the deviation. Furthermore, according to this embodiment, a worn area capture image is generated for each of the plurality of image processing execution programs M, and an evaluation value is calculated for each of these worn area capture images. The worn area capture image with the highest evaluation value is selected, and both the residual diameter and the displacement of the contact line 9 are calculated based on this image. Specifically, a number of worn area capture images corresponding to the number of image processing execution programs M are generated, and the most suitable worn area capture image for residual diameter and displacement evaluation is selected as the target for residual diameter and displacement evaluation. As a result, the processing result of the image processing execution program M that best processes the linear sensor image 50 being processed is selected from the plurality of registered image processing execution programs M and used as the worn area capture image. In this way, since the operator does not need to select a combination of image processing functions suitable for the linear sensor image 50 to be processed on a trial and error basis, and a wear area capture image suitable for calculating the residual diameter and deviation of the contact line 9 can be generated, and the residual diameter and deviation can be measured, the operator's time for selecting the image processing function suitable for the input image can be reduced. Furthermore, according to this embodiment, priorities are stored in association with each of the plurality of image processing execution programs M. When the same evaluation value is calculated for a plurality of worn area captured images, the worn area captured image evaluating unit 34 selects the worn area captured image corresponding to the image processing execution program M with the higher priority as the worn area captured image with the highest evaluation value. Therefore, when the same evaluation value is calculated for a plurality of worn area captured images, the worn area captured image corresponding to the image processing execution program M with the higher priority is selected as the worn area captured image with the highest evaluation value. Therefore, even when there are a plurality of worn area captured images with the same evaluation value, the worn area captured image corresponding to the image processing execution program M with the higher priority can be selected, making it easier to select a worn area captured image suitable for residual diameter and displacement evaluation. Furthermore, according to this embodiment, the worn area captured image evaluation unit 34 calculates the number of worn areas 9a for each of the plurality of worn area captured images, and calculates the evaluation value so that the evaluation value becomes higher when the difference between the number of worn areas 9a and the actual number of contact lines 9 is smaller. Therefore, it is easy to select a worn area captured image in which the number of worn areas 9a matches the actual number of contact lines 9 as the worn area captured image for calculating the residual diameter and the offset. Consequently, the accuracy of capturing the worn areas 9a of the contact lines 9 can be improved, and the accuracy of calculating the residual diameter and the offset can also be improved. Furthermore, according to this embodiment, by selecting an appropriate image processing program based on the environment and generating a corresponding captured image of the worn area, the accuracy of residual diameter and displacement calculations can be improved, while also reducing processing time and processing load. Furthermore, by not executing image processing programs M that are inappropriate for the selected environment, the processing time and memory required to execute the contact line detection device 20 can be reduced. In addition, according to this embodiment, since the wear area capture image with the highest evaluation value, i.e., the contact line likelihood, is selected, it is easy to select the wear area capture image whose features on the image of the wear area 9a are similar to the actual appearance of the contact line 9 as the wear area capture image for calculating the residual diameter and the deviation. Therefore, the capture accuracy of the wear area 9a of the contact line 9 can be improved, and the calculation accuracy of the residual diameter and the deviation can also be improved. Furthermore, in the above-mentioned embodiment, for example, for the sake of convenience, the post-processing unit 41 is described as being included in the wear area capture image generation unit 32, but the present invention is not limited thereto. The post-processing unit 41 may also be included in the wear area capture image evaluation unit 34, or may be independent of the wear area capture image generation unit 32 and the wear area capture image evaluation unit 34 as independent functional structures. Furthermore, as long as it does not deviate from the essence of the present invention, the structure within the control device 23 may of course be changed. Furthermore, in the above-mentioned embodiment, the contact line detection device includes both the contact line deviation calculation unit 35 and the contact line residual diameter calculation unit 36, and is configured to calculate both the residual diameter and deviation of the contact line 9, but the present invention is not limited thereto. For example, the contact line detection device may also be configured to calculate only one of the residual diameter and deviation of the contact line. In addition, the present invention is not limited to the above-mentioned embodiments and various modifications described with reference to the drawings. As long as it does not depart from the essence of the present invention, the structures listed in the above-mentioned embodiments and various modifications can be selected or replaced, or appropriately changed to other structures. 1: Inspection Vehicle 1a: Top 1b: Current Collector 2: Line 3: Pole 4: Feeder 5: Crossbeam 6: Bent Traction Metal 7: Hanging Wire 8: Hanger 9: Contact Wire 9a: Worn Area 9b: Surface 9c, 9d: Dark Area 10: Blank Area 20: Contact Wire Detection Device 21: Illumination 22: Linear Sensor 22a: Scan Line 22b: Direction 22c: Sensing Surface 23: Control Device 30: Linear Sensor Image Generator 31: Environment Determination Unit 32: Worn Area Capture Image Generator 33: Image Processing Execution Program Memory 33a: Processing Storage Unit 33b: Parameter Group Storage Unit 33c: Processing Mode Registration Unit 34: Worn Area Capture Image Evaluation Unit 35: Contact Wire Deviation Calculator (Contact Wire Detection Unit) 36: Contact Wire Remaining Diameter Calculator (Contact Wire Detection Unit) 40: Image processing unit 40A: Contrast adjustment processing unit 40B: Background / noise removal processing unit 40C: Binarization processing unit 40D: Interpolation processing unit 40E: Wear area correction unit 41: Post-processing unit 50: Linear sensor images 50A, 50B: Wear area capture image D: Offset Dr: Residual diameter d: Center of gravity coordinates f: Focal length h, hs: Distance M, M1, M2, M3: Image processing execution program M1~ML: Processing mode P1~PZ: Image processing functions PR1~PRX: Image processing functions PS1~PSY: Parameter group PS2: Parameter group p: Number of pixels R: Extension direction s: Length w, W: Wear surface width X: Horizontal Y: Vertical FIG1 is an explanatory diagram showing the driving environment of an electric railway vehicle equipped with a contact wire detection device according to an embodiment of the present invention. FIG2 is a block diagram showing the functional structure of the contact wire detection device according to this embodiment. FIG3 is a schematic diagram showing an example of a linear sensor image generated based on an image captured by the linear sensor of the contact wire detection device according to this embodiment. FIG4 is a schematic diagram showing an example of a wear area capture image temporarily obtained by the contact wire detection device according to this embodiment. FIG5 is a schematic diagram showing an example of a wear area capture image finally obtained by the contact wire detection device according to this embodiment. FIG6 is an explanatory diagram of the image processing execution program memory unit of the contact wire detection device according to this embodiment. FIG7(a) is an explanatory diagram of the contact wire deviation calculation unit of the above-mentioned contact wire detection device, and FIG7(b) is an explanatory diagram of the contact wire residual diameter calculation unit of the above-mentioned contact wire detection device. FIG8 is a flowchart for explaining the operation of the above-mentioned contact wire detection method performed by the contact wire detection device according to this embodiment. 20: Contact line detection device 21:Lighting 22: Linear sensor 23: Control device 30: Linear sensor image generation unit 31: Environmental Assessment Department 32: Wear area capture image generation unit 33: Image processing execution program memory unit 34: Wear area capture image evaluation unit 35: Contact line deviation calculation unit (contact line detection unit) 36: Contact wire residual diameter calculation unit (contact wire detection unit) 40: Image processing unit 40A: Contrast adjustment processing unit 40B: Background / noise removal processing unit 40C: Binarization processing unit 40D: Interpolation processing unit 40E: Wear area correction part 41: Post-processing unit

Claims

1. A contact wire detection device comprising a linear sensor for capturing images of a contact wire, and using an image captured by the linear sensor as an input image to detect the contact wire; characterized in that it comprises: a linear sensor image generation unit that arranges the input images in a time sequence to generate a linear sensor image; an image processing execution program memory unit that, when sequentially executing a plurality of image processing functions on the linear sensor image, registers a plurality of execution orders as image processing execution programs; and a wear region capturing image generation unit that, according to each of the plurality of image processing execution programs, uses the linear sensor image as input to execute the plurality of image processing functions to capture a wear region of the contact wire, and by scanning outward from the captured wear region of the contact wire, using a position with a brightness value below a preset brightness value as the boundary of the wear region, corrects the wear region, and generates a plurality of wear region capturing images corresponding to each of the plurality of image processing execution programs. The wear area image evaluation unit calculates the evaluation value of each of the plurality of wear area images and selects the wear area image with the highest evaluation value; and the contact line detection unit calculates either or both of the residual diameter and offset of the contact line based on the selected wear area image with the highest evaluation value.

2. The contact wire detection device as described in claim 1, wherein, For each of the aforementioned plurality of image processing execution programs, the following are executed sequentially: contrast adjustment processing (0 or more), background / noise removal processing (0 or more), binarization processing (1 or more), and interpolation processing (0 or more).

3. The contact wire detection device as requested in item 1 or 2, wherein, Each of the plurality of image processing execution programs is assigned a priority, and when the same evaluation value is calculated for the plurality of wear region captured images, the wear region captured image evaluation unit selects the wear region captured image corresponding to the image processing execution program with the highest priority as the wear region captured image with the highest evaluation value.

4. The contact wire detection device as described in claim 3, wherein, When calculating the number of wear regions for each of the plurality of images captured from wear regions, and the difference between the number of wear regions and the actual number of contact lines is small, the wear region image evaluation unit calculates the evaluation value in a manner that increases the evaluation value.

5. The contact wire detection device as described in claim 3, wherein, The aforementioned wear area image evaluation unit calculates the contact line likelihood for each of the plurality of wear area images, and calculates the evaluation value by increasing the evaluation value when the contact line likelihood is high. The contact line likelihood is an index that indicates that the wear area has characteristics that approximate the contact line in the image.

6. The contact wire detection device as described in claim 1, wherein, The image processing execution program memory unit is equipped with an image processing execution program corresponding to each of the plurality of different environments in which the image is captured by the linear sensor. It further includes an environment determination unit that determines the environment in which the input image is captured based on the linear sensor image. The wear area capture image generation unit generates the wear area capture image according to the image processing execution program corresponding to the determined environment.

7. A contact wire detection method, which uses an image captured by a linear sensor for capturing the contact wire as an input image and detects the contact wire accordingly; characterized in that it includes the following steps: arranging the input image in a time sequence to generate a linear sensor image; when sequentially executing a plurality of image processing functions on the linear sensor image, registering a plurality of execution orders as image processing execution programs, and executing the plurality of image processing functions with the linear sensor image as input according to each of the plurality of image processing execution programs, thereby capturing the wear area of ​​the contact wire; scanning outward from the captured wear area of ​​the contact wire, and using a position with a brightness value below a preset brightness value as the boundary of the wear area, thereby correcting the wear area and generating a plurality of wear area capture images corresponding to each of the plurality of image processing execution programs; Calculate the evaluation value of each of the plurality of wear area captured images, select the wear area captured image with the highest evaluation value; and calculate either or both of the residual diameter and offset of the contact line based on the selected wear area captured image with the highest evaluation value.

Images