Shape detection system

Abstract

To provide a system that can quickly grasp unevenness on surfaces of a working face and rock mass.SOLUTION: A shape detection system comprises: a plurality of light sources 2a, 2b, and 2c that irradiates a working surface 1 with light of mutually different color tones; a camera 3 for imaging the working surface 1; and a computer 4 for analyzing images captured by the camera 3. The light sources 2a, 2b, and 2c are arranged so as to be spaced apart from each other and irradiate the working surface 1 with light from mutually different angles relative to the working surface 1. The computer 4 detects the color tones of each pixel in the image captured by the camera 3 and calculates an inclination angle and inclination direction of the unevenness of the working surface 1 on the basis of the detected color tones.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a shape detection system. [Background technology]

[0002] When excavating a tunnel, it has been customary to measure the strike and inclination of the tunnel face to understand the condition of the tunnel face and the surrounding ground. By measuring the strike and inclination of the tunnel face in this way, it is possible to evaluate the stability of the ground, such as the possibility of rock collapse.

[0003] Patent Document 1 discloses a method for evaluating the stability of the ground by creating a TIN model from three-dimensional information on the tunnel face obtained by a three-dimensional laser scanner, and then creating an estimated map of the ground ahead of the tunnel based on this TIN model. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2013-72704 Summary of the Invention [Problem to be solved by the invention]

[0005] In the invention disclosed in Patent Document 1, the strike and dip of each triangular face of a TIN model are calculated to determine the distribution of the strike and dip of the triangular faces, and from the distribution of the strike and dip of these triangular faces, discontinuous faces that differ in strike and dip from the excavation face, which is the face when the tunnel face is considered as a single face, are extracted to create an estimated map of the natural ground ahead of the tunnel.This type of analysis method using laser surveying requires a lot of time for analysis, so it is not appropriate for application to construction work where the time interval between excavation processes is short.

[0006] The present invention aims to provide a system that can quickly grasp the unevenness of the face and rock surface. [Means for solving the problem]

[0007] The present invention is a shape detection system for detecting the shape of a working face or bedrock surface, and comprises a plurality of light sources that irradiate light of different colors toward the surface, an imaging unit for imaging the surface, and an image analysis unit that analyzes the image captured by the imaging unit, wherein the plurality of light sources are positioned so as to be spaced apart from one another and irradiate light toward the surface from different angles relative to the surface, and the image analysis unit detects the color tone at each pixel of the image captured by the imaging unit, and calculates the inclination angle and inclination direction of the surface irregularities based on the detected color tone. [Effects of the Invention]

[0008] According to the present invention, irregularities on the surface of the working face or rock mass can be detected quickly. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a cross-sectional view of the vicinity of a tunnel face in the longitudinal direction of a tunnel to which a shape detection system according to an embodiment of the present invention is applied. [Figure 2] 1 is a front view of a tunnel face to which a shape detection system according to an embodiment of the present invention is applied. [Figure 3] This is a front view of the tunnel face captured by a camera. [Figure 4] 10 is a flowchart showing the flow of work in a preparation stage for detecting unevenness by the shape detection system according to the embodiment of the present invention. [Figure 5] FIG. 1 is a front view of a model body according to an embodiment of the present invention. [Figure 6] 10 is a graph showing the relationship between RGB values and tilt direction (angle τ) when the tilt angle θ is 20°. [Figure 7] 10 is a graph showing the relationship between RGB values and tilt direction (angle τ) when the tilt angle θ=60°. [Figure 8]10 is a flowchart showing a procedure for detecting irregularities using the shape detection system according to the embodiment of the present invention. [Figure 9] 10 is a flowchart showing a procedure for detecting irregularities using the shape detection system according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0010] A shape detection system 100 according to an embodiment of the present invention will be described below with reference to the drawings.

[0011] The shape detection system 100 of this embodiment is used to grasp the unevenness of the surface of a tunnel face or bedrock. Note that the following embodiment will be described taking as an example a case where the shape detection system 100 is used to detect the surface shape of a face 1 of a tunnel T, but the shape detection system 100 can also be applied to detecting the surface shape of a bedrock in civil engineering works such as dam construction, cut earth construction, and bridge pier foundations.

[0012] First, the configuration of the shape detection system 100 will be described with reference to Figures 1 and 2. Figure 1 is a cross-sectional view of the vicinity of a tunnel face 1 in the longitudinal direction of a tunnel T. Figure 2 is a front view of the tunnel face 1 of the tunnel T.

[0013] As shown in Figures 1 and 2, the shape detection system 100 includes a plurality of light sources 2 that irradiate the working face 1 (surface), a camera 3 (imaging unit) for imaging the working face 1, and a computer 4 (image analysis unit) that analyzes the image captured by the camera 3.

[0014] In this embodiment, three light sources 2 are used: light source 2a that emits blue light, light source 2b that emits green light, and light source 2c that emits red light. Light sources 2a, 2b, and 2c are spaced apart near the inner wall of tunnel T and are arranged so as to irradiate light toward the center of tunnel T from different angles relative to the tunnel face 1. In this embodiment, as shown in Fig. 2, light source 2b is provided near the top end of tunnel T, and light sources 2a and 2c are provided near the inner wall at the lower end (ground) of tunnel T.

[0015] The camera 3 is placed in a position where it can capture an image of the entire face 1 of the tunnel T. The image captured by the camera 3 is sent to the computer 4, where it is processed and stored.

[0016] Next, a method for detecting irregularities (surface shape) of the working face 1 using the shape detection system 100 will be described.

[0017] 3, the surface of the working face 1 is uneven, and cracks may be present on the surface. Therefore, in the shape detection system 100, three colors of light are irradiated onto the working face 1 from light sources 2a, 2b, and 2c, and the color tone of the surface of the working face 1 is analyzed by a camera 3 and a computer 4, thereby detecting the uneven shape of the working face 1 and the direction of the cracks.

[0018] First, as a preparation step, the relationship between the RGB values (color tone), inclination angle θ, and inclination direction (angle τ) of the face 1 is obtained in order to determine the unevenness of the face 1. The inclination angle θ refers to the inclination angle of the inclined planes that constitute the unevenness on the surface of the face 1 and the inclined planes of a model body 10, which will be described later. The inclination direction refers to the inclination direction of the inclined planes that constitute the unevenness and the inclination direction of the inclined planes of a model body 10, which will be described later. For cracks that exist on the surface of the face 1, the direction in which the crack extends is referred to as the inclination direction. In this embodiment, the inclination direction is represented by the angle τ.

[0019] A method for obtaining the relationship between RGB values, tilt angle θ, and tilt direction (angle τ) will be specifically described below with reference to Figures 4 to 7. First, the flow of work in the preparation stage will be described with reference to the flowchart shown in Figure 4.

[0020] In step S11, the model body 10 is imaged under normal lighting. Specifically, the model body 10 shown in Fig. 5 is placed at the lower end of the working face 1 and near the center in the width direction (for example, the area surrounded by dotted line F in Fig. 2), and so that the reference plane (section S1) is perpendicular to the longitudinal axis of the tunnel T. Note that in this embodiment, for convenience of explanation, the model body 10 is placed in the orientation shown in Fig. 5, specifically, so that region R3 is located above and region R7 is located below.

[0021] Next, the light sources 2a, 2b, and 2c are not turned on, and the model body 10 is imaged by the camera 3 with the lighting normally used in the tunnel T turned on. Note that "under normal lighting" here means a state in which the lighting normally used in the tunnel T is turned on. Here, the shape of the model body 10 will be described.

[0022] 5, the model body 10 of this embodiment is a regular octagonal plate-like member. The model body 10 is divided into four sections, S1 to S4, in order from the inside in the radial direction. The model body 10 is also divided into eight regions R1 to R8 in the circumferential direction. Each of the regions R1 to R8 is formed by an area surrounded by two straight lines extending from adjacent vertices of the octagon toward the center O and the sides connecting the adjacent vertices.

[0023] Region R1 has three sections S21, S31, and S41 corresponding to sections S2 to S4. Similarly, region R2 has sections S22, S32, and S42, region R3 has sections S23, S33, and S43, region R4 has sections S24, S34, and S44, region R5 has sections S25, S35, and S45, region R6 has sections S26, S36, and S46, region R7 has sections S27, S37, and S47, and region R8 has sections S28, S38, and S48.

[0024] The section S1 is formed by a regular octagonal plane. The plane that forms the section S1 forms a reference plane (with an inclination angle θ=0).

[0025] The sections S1, S2, and S3 protrude from the reference plane and have a regular octagonal outer shape.

[0026] Section S2 is formed so as to surround the outer periphery of section S1. Sections S21 to S28 are formed by dividing section S2 into eight equal sections in the circumferential direction. Each of sections S21 to S28 has a flat surface (inclined surface) on the front side that is inclined at an angle of 20° toward the center O (reference plane) of the model body 10.

[0027] Section S3 is formed to surround the outer periphery of section S2. Sections S31 to S38 are formed by dividing section S3 into eight equal sections in the circumferential direction. Each of sections S31 to S23 has a flat surface (inclined surface) on the front side that is inclined at an angle of 40° toward the center O (reference plane) of the model body 10.

[0028] Section S4 is formed so as to surround the outer periphery of section S4. Sections S31 to S38 are formed by dividing section S3 into eight equal sections in the circumferential direction. Each of sections S31 to S38 has a flat surface (inclined surface) on the front side that is inclined at an angle of 60° toward the center O (reference plane) of the model body 10.

[0029] In the following, since the sections S2, S3, and S4 differ only in the inclination angle of the inclined surface, only the section S2 will be described, and a description of the sections S3 and S4 will be omitted.

[0030] 5, when the reference plane has an inclination angle θ=0, sections S21 to S28 in model body 10 each have an inclination angle θ=20°. Furthermore, when the inclination direction of the inclined surface of section S21 in model body 10 is set to angle τ=0 (zero)°, the inclined surface of section S22 has an angle τ=45°, the inclined surface of section S23 has an angle τ=90°, the inclined surface of section S24 has an angle τ=135°, the inclined surface of section S25 has an angle τ=180°, the inclined surface of section S26 has an angle τ=225°, the inclined surface of section S27 has an angle τ=270°, and the inclined surface of section S28 has an angle τ=315°.

[0031] The planes (outer walls) that form the peripheries of the sections S2 to S4 are formed by planes that are perpendicular to the reference plane (the plane in the section S1) (see cross sections AA and BB in FIG. 5).

[0032] 5 shows an example of the model body 10. The model body 10 is not limited to an octagonal shape, but may have any polygonal shape, such as a hexagon or a decagon. The model body 10 is not limited to a polygonal shape, but may also be circular. The radial section may also be divided into four or more sections, that is, five or more sections with different inclination angles of the inclined surfaces.

[0033] The model body 10 is positioned as described above, and an image P1 of the model body 10 is captured with the lights installed in the tunnel T turned on without irradiating light from the light sources 2a, 2b, and 2c. The image P1 is sent to the computer 4 and stored in the computer 4.

[0034] In step S12, the model body 10 is imaged under illumination by three color light sources (light sources 2a, 2b, and 2c). Specifically, the model body 10 is imaged by irradiating it with light from the light sources 2a, 2b, and 2c. At this time, as shown in FIG. 2, the light source 2a emitting blue light is positioned on an extension of the model body 10 at an angle τ=0° (region R1), the light source 2b emitting green light is positioned on an extension of the model body 10 at an angle τ=90° (region R3), and the light source 2c emitting red light is positioned on an extension of the model body 10 at an angle τ=180° (region R5). When imaging the model body 10, the lighting installed in the tunnel T is kept on. The image P2 of the model body 10 captured by the camera 3 is sent to the computer 4 and stored in the computer 4.

[0035] In step S13, the amount of change in RGB values is calculated. Specifically, the computer 4 calculates the amount of change in RGB values (change in color tone) between image P1 of the model body 10 captured in step S11 and image P2 of the model body 10 captured in step S12. More specifically, the computer 4 calculates the difference (amount of change) in RGB values for each pixel between image P1 and image P2.

[0036] Then, for each partition, the computer 4 calculates the average value of the differences in RGB values of all pixels in the partition and stores each calculated average value as an RGB reference value (reference value) that serves as the basis for determining each partition. Note that if it is expected that there will be areas that will be shaded on the slope of each partition, an area that will not be shaded within the partition may be set in advance, and the average value of the differences in RGB values of each pixel in this area may be used as the RGB reference value (reference value).

[0037] In step S14, the relationship between the RGB values, the tilt angle θ, and the tilt direction (angle τ) is calculated. A method for calculating the relationship between the RGB values, the tilt angle θ, and the tilt direction (angle τ) will be specifically described below.

[0038] As described above, the sections S1 to S4 each have a different inclination angle θ, and the regions R1 to R8 each have a different inclination direction (angle τ). Therefore, when the model 10 is irradiated with light from the light sources 2a to 2c, the RGB values differ depending on the section.

[0039] In this embodiment, as described above, the light source 2a emitting blue light is positioned on an extension of the model body 10 at an angle τ=0° (region R1), the light source 2b emitting green light is positioned on an extension of the model body 10 at an angle τ=90° (region R3), and the light source 2c emitting red light is positioned on an extension of the model body 10 at an angle τ=180° (region R5).

[0040] The inclined surfaces of sections S21, S31, and S41 in region R1 (angle τ = 0°) are located on the rear side when viewed from the light source 2a, so blue light from the light source 2a is unlikely to hit the inclined surfaces of sections S21, S31, and S41. Therefore, the B value in the RGB values of region R1 is smaller (smallest) than the B values of the other regions R2 to R8. Furthermore, as the inclination angle θ of the inclined surface increases, it becomes more difficult for blue light to hit the inclined surface, so the B value decreases in the order of sections S21, S31, and S41. In section S41, the inclination angle θ of the inclined surface is the largest (60°), so almost no blue light from the light source 2a hits the inclined surfaces. Therefore, the B value of section S41 is 0 (zero).

[0041] Similarly, the inclined surfaces of sections S23, S33, and S43 in region R3 (angle τ = 90°) are located on the rear side when viewed from light source 2b, and therefore green light from light source 2b is unlikely to strike the inclined surfaces of sections S23, S33, and S43. Therefore, the G value in the RGB values of region R3 is smaller (smallest) than the B value in the other regions R1, R2, and R4 to R8. Furthermore, as the inclination angle θ of the inclined surface increases, it becomes more difficult for green light to strike the inclined surface, so the G value decreases in the order of sections S23, S33, and S43. In section S43, the inclination angle θ of the inclined surface is the largest (60°), and therefore almost no green light from light source 2b strikes the inclined surface. Therefore, the G value of section S43 is 0 (zero).

[0042] Furthermore, since the inclined surfaces of sections S25, S35, and S45 in region R5 (angle τ=180°) are located on the rear side when viewed from light source 2c, red light from light source 2c is unlikely to strike the inclined surfaces of sections S25, S35, and S45. Therefore, the R value in the RGB values of region R5 is smaller (smallest) than the R values of the other regions R1 to R4 and R6 to R8. Furthermore, as the inclination angle θ of the inclined surface increases, it becomes more difficult for red light to strike the inclined surface, so the R value decreases in the order of sections S25, S35, and S45. In section S45, the inclination angle θ of the inclined surface is the largest (60°), so almost no red light from light source 2c strikes the inclined surface. Therefore, the R value of section S45 is 0 (zero).

[0043] Conversely, region R1 (angle τ = 0°) is most susceptible to red light from light source 2c, and therefore the R value in the RGB values is greater (maximum) than the R values in regions R2 to R8. The R values in each of regions R1 to R8 decrease from region R1 to region R5 due to the relationship of the tilt direction (angle τ), specifically, from region R1 to region R5, in the order of region R2, region R3, region R4, and region R5, and from region R1 to region R8, region R7, region R6, and region R5. Furthermore, the greater the tilt angle θ of the inclined surface, the more susceptible red light is to the inclined surface, and therefore the R values increase in the order of sections S21, S31, and S41. In section S43, the tilt angle θ of the inclined surface is the largest (60°), and therefore the red light from light source 2c hits the inclined surface the most. Therefore, the R value is greatest in section S41.

[0044] Similarly, region R5 (angle τ=180°) is most susceptible to blue light from light source 2a, and therefore the B value in the RGB values is greater (maximum) in region R5 (angle τ=180°) than the B values in regions R1 to R4 and R6 to R8. Note that, due to the relationship of the tilt direction (angle τ), the B values in each of regions R1 to R8 decrease from region R5 toward region R1, specifically, from region R5 to region R4, region R3, region R2, and region R1, and from region R5 to region R6, region R7, region R8, and region R1. Furthermore, the larger the tilt angle θ of the inclined surface, the more susceptible blue light is to the inclined surface, and therefore the B values increase in the order of sections S25, S35, and S45. In section S45, the tilt angle θ of the inclined surface is greatest (60°), and therefore the blue light from light source 2a hits the inclined surface the most. Therefore, the B value is greatest in section S45.

[0045] Furthermore, because region R7 (angle τ=270°) is most susceptible to green light from light source 2b, the G value in the RGB values for region R7 (angle τ=270°) is greater (maximum) than the G values for the other regions R1 to R6 and R8. The G values for each of regions R1 to R8 decrease from region R7 to region R3 due to the relationship between the tilt direction (angle τ), specifically, from region R7 to region R8, region R1, region R2, and region R3, and from region R7 to region R6, region R5, region R4, and region R3. Furthermore, because the greater the tilt angle θ of the inclined surface, the more susceptible green light is to the inclined surface, the greater the G value becomes in the order of sections S27, S37, and S47. In section S47, the tilt angle θ of the inclined surface is the largest (60°), so the green light from light source 2b hits the inclined surface the most. Therefore, the G value is greatest in section S47.

[0046] The relationship between the RGB reference value, tilt angle θ, and tilt direction (angle τ) for each section calculated in step S13 can be plotted on graphs as shown in Figures 6 and 7. Note that Figure 6 is a graph showing the relationship between the RGB reference value and angle τ when the tilt angle θ=20°, and Figure 7 is a graph showing the relationship between the RGB reference value and angle τ when the tilt angle θ=60°. Note that the tilt angle θ=40° is not shown. The vertical axis in Figures 6 and 7 represents luminance (the magnitude of the RGB value).

[0047] 6 and 7, each color will form an approximately sinusoidal curve (see the dotted lines in FIGS. 6 and 7). The computer 4 stores the relationship between the brightness (RGB reference values) and the tilt direction (angle τ) shown by this sinusoidal curve for each tilt angle θ = 20, 40, and 60.

[0048] Next, a method for detecting irregularities on the working face 1 will be described with reference to FIG.

[0049] In step S21, the tunnel face 1 is imaged under normal lighting. Specifically, the tunnel face 1 is imaged by the camera 3 with the lighting installed in the tunnel T turned on, without irradiating it with light from the light sources 2a, 2b, and 2c. The image P3 of the tunnel face 1 imaged by the camera 3 is sent to the computer 4 and stored in the computer 4.

[0050] In step S22, the tunnel face 1 is imaged under illumination by the three-color light sources (light sources 2a, 2b, and 2c). Specifically, the light from the light sources 2a, 2b, and 2c is irradiated onto the tunnel face 1 to image the tunnel face 1. At this time, the lighting installed inside the tunnel T remains on. The image P4 of the tunnel face 1 imaged by the camera 3 is sent to the computer 4 and stored therein.

[0051] In step S23, the amount of change in RGB values is calculated. Specifically, the amount of change in RGB values (change in color tone) between image P3 of the working face 1 captured in step S21 and image P4 of the working face 1 captured in step S22 is calculated. More specifically, the difference in RGB values for each pixel between image P3 and image P4 is calculated.

[0052] In step S24, the inclination angle θr and inclination direction (angle τr) are calculated for each pixel of the working face 1. A method for calculating the values of the inclination angle θr and inclination direction (angle τr) for each pixel in the image of the working face 1 will be specifically described below with reference to the flowchart shown in FIG.

[0053] In step S241, the magnitude relationship of the amount of change in the RGB values (hereinafter also referred to as "RGB measurement values") obtained in step S23 is determined. Specifically, the computer 4 determines the maximum value of the RGB measurement values and its color, the intermediate value and its color, and the minimum value and its color.

[0054] In step S242, the absolute value D1 of the difference between the maximum value and the median value, and the absolute value D2 of the difference between the median value and the minimum value are calculated.

[0055] In step S243, the ratio Rd between the absolute value D1 and the absolute value D2 is calculated.

[0056] In step S244, the tilt direction (angle τr) is calculated. Specifically, first, the range of angle τ is limited based on the magnitude relationship of the RGB measurement values calculated in step S241. For example, if the maximum value of the RGB measurement values is the G value, the intermediate value is the R value, and the minimum value is the B value, it can be seen from FIGS. 6 and 7 that the range of angle τr to be calculated is angle τ = 270° to 315°. Then, computer 4 calculates angle τr that satisfies ratio Rd (= D1 / D2) within the angle τ range of 270° to 315°.

[0057] In step S245, the tilt angle θr is calculated. Specifically, the tilt angle θr is calculated based on the absolute values D1 and D2 calculated in step S242 and the angle τr calculated in step S244. More specifically, for tilt angles θ=20°, 40°, and 60°, the tilt angle θ at which the absolute values D1 and D2 match at the angle τr calculated in step S244 is extracted as the tilt angle θr.

[0058] In this embodiment, since the ratio Rd between the absolute value D1 and the absolute value D2 at the same angle τ is approximately the same for inclination angles θ=20°, 40°, and 60°, the angle τr is calculated in steps S224 and S245, and then the inclination angle θr is calculated based on this angle τr. However, for example, the absolute values D1 and D2 calculated in step S242 may be directly compared with the RGB reference values for inclination angles θ=20°, 40°, and 60° to simultaneously calculate the inclination angle θr and inclination direction (angle τr) of the irregularities on the working face 1.

[0059] By calculating the inclination angle θr and inclination direction (angle τr) of each pixel in this way, it is possible to detect the inclination angle θr and inclination direction (angle τr) at the position of the working face 1 corresponding to each pixel. The calculated inclination angle θr and angle τr can be displayed, for example, as a histogram, allowing the tendency of the inclination angle θr and inclination direction (angle τr) of the unevenness of the working face 1 to be visually recognized.

[0060] Note that cracks present on the working face 1 do not reflect light from the three-color light sources (light sources 2a, 2b, 2c), so they appear close to black, with all RGB values being small. Therefore, pixels for which all RGB values are below a predetermined value are assigned a flag or the like corresponding to a crack, without determining the tilt angle θr and tilt direction (angle τr). This makes it easy to recognize the presence of cracks on the working face 1.

[0061] The inclination angle θr and the angle τr calculated as described above may be used to generate a three-dimensional image of the working face 1. In this case, the positions and shapes of irregularities and cracks on the working face 1 can be visually recognized easily.

[0062] It should be noted that the work in the preparation stage (steps S11 to S14) may not be performed depending on the situation. For example, when the work of detecting unevenness on the face 1 is repeatedly performed in accordance with the progress of the excavation work of the tunnel T, it is considered that the influence of the lighting inside the tunnel T will hardly change, so once the work of calculating the RGB reference values shown in steps S11 to S14 in Fig. 4 (the work of calculating the relationship between the inclination angle θ, brightness, and inclination direction (angle τ) shown in Fig. 6 and Fig. 7) is performed, the subsequent steps may be omitted.

[0063] Furthermore, if the influence of the lighting inside the tunnel T is small (for example, if the lighting in the tunnel T can be turned off or if the area near the face 1 is dark), the RGB values may be calculated from the image P2 of the model body 10 captured under the illumination of the three-color light source (light sources 2a, 2b, and 2c), and this RGB value may be set as the RGB reference value as is. In this case, the operations of steps S11 and S13 in FIG. 4 can be omitted.

[0064] In the above embodiment, the three primary colors of red, green, and blue are used as the light source 2, but the present invention is not limited to this and any light source may be used as long as the light sources 2a, 2b, and 2c emit light of different color tones (wavelengths). In this case, not only visible light but also ultraviolet light, infrared light, etc. may be used.

[0065] In the above embodiment, the light source 2b arranged at the top end emits green light, but it is possible to appropriately select which color the light sources 2a, 2b, and 2c emit, from red, blue, and green.

[0066] In addition, in the above embodiment, the case where the face 1 is semicircular is exemplified, but if the face 1 is circular, for example, the light sources 2a, 2b, and 2c may be arranged at intervals of 120° along the inner wall of the tunnel T.

[0067] The above-described shape detection system 100 provides the following advantages.

[0068] The shape detection system 100 detects the color tone (RGB value) of each pixel of the image P4 captured by the camera 3, and calculates the inclination angle θr and inclination direction (angle τr) of the unevenness on the working face 1 based on the detected color tone (RGB value). This means that the amount of data processing required is smaller than, for example, analysis by laser surveying, and unevenness present on the working face 1 can be detected quickly.

[0069] Furthermore, the shape detection system 100 can determine whether or not a crack exists on the working face 1 from the image P4 captured by the camera 3. Furthermore, when a crack exists, the shape detection system 100 can predict the risk of collapse of the working face 1 based on the size and direction of the crack itself, and the inclination angle θr and inclination direction (angle τr) of the irregularities around the crack.

[0070] In the shape detection system 100, RGB measurement values are detected based on the difference between an image P3 of the working face 1 captured without irradiating it with light from the light sources 2a, 2b, and 2c, and an image P4 of the working face 1 captured with irradiating it with light from the light sources 2a, 2b, and 2c. For example, if the working face 1 is a rock mass with large color variations (large color variations depending on the geological layer), the color of the rock mass itself may be added to the RGB measurement values. Therefore, by using the difference between the image P3 of the working face 1 captured without irradiating it with light from the light sources 2a, 2b, and 2c and the image P4 of the working face 1 captured with irradiating it with light from the light sources 2a, 2b, and 2c as the RGB measurement value, i.e., the amount of change in RGB values due to the presence or absence of light from the light sources 2a, 2b, and 2c, the influence of the color of the rock mass itself on the RGB measurement values can be eliminated.

[0071] Furthermore, the shape detection system 100 detects the RGB reference value based on the difference between an image P1 of the model body 10 captured without irradiating it with light from the light sources 2a, 2b, and 2c, and an image P2 of the model body 10 captured with irradiating it with light from the light sources 2a, 2b, and 2c. In this way, by using the difference between the images P1 and P2, that is, the amount of change in the RGB values due to the presence or absence of light from the light sources 2a, 2b, and 2c, as the judgment standard (RGB reference value), it is possible to eliminate the influence of lighting installed inside the tunnel T.

[0072] Although the embodiments of the present invention have been described above, the above embodiments merely illustrate some of the application examples of the present invention, and it is not intended that the technical scope of the present invention be limited to the specific configurations of the above embodiments.

[0073] In the above embodiment, RGB values are detected as the color tones of the pixels of the images P1 to P4, but this is not limiting. For example, the L*a*b* color space of the pixels of the images P1 to P4 may be detected.

[0074] Furthermore, in the above embodiment, three light sources 2 are used, but two, or four or more may be used as long as they emit light of different color tones.

[0075] In the above embodiment, the ratio Rd is calculated based on the absolute value D1 of the difference between the maximum value and the median value, and the absolute value D2 of the difference between the median value and the minimum value. However, the present invention is not limited to this, and the ratio Rd may be calculated based on the absolute value of the difference between any two RGB values and the absolute value of the difference between any two values different from the any two values. [Explanation of symbols]

[0076] 100···Shape detection system 1. Face 2...Light source 2a...Light source 2b...Light source 2c...Light source 3. Camera (imaging unit) 4. Computer (image analysis section) 10. Model body

Claims

1. A shape detection system for detecting the shape of a working face or a rock surface, a plurality of light sources that emit light of different colors toward the surface; an imaging unit for imaging the surface; an image analysis unit that analyzes the image captured by the imaging unit, the plurality of light sources are spaced apart from one another and arranged to irradiate light toward the surface from different angles relative to the surface; The image analysis unit Detecting a color tone for each pixel of an image captured by the imaging unit; A shape detection system that calculates the inclination angle and inclination direction of the surface irregularities based on the detected color tone.

2. 2. The shape detection system according to claim 1, A shape detection system that detects the color tone based on the difference between an image of the surface captured without irradiating it with light from the multiple light sources and an image of the surface captured with irradiating it with light from the multiple light sources.

3. 2. The shape detection system according to claim 1, Further provided is a model body having a plurality of planes with different inclination angles and inclination directions, A shape detection system that sets a reference value for the color tone based on the difference between an image captured without irradiating the model body placed near the surface with light from the multiple light sources and an image captured after irradiating the model body placed near the surface with light from the multiple light sources, and detects the color tone of the surface based on the reference value.

4. 4. A shape detection system according to claim 1, the plurality of light sources are three light sources that respectively emit red, green, and blue light; A shape detection system in which the color tone is a value based on RGB values.

5. 5. The shape detection system according to claim 4, A shape detection system that calculates the tilt angle based on the absolute value of the difference between any two of the RGB values and the absolute value of the difference between any two values different from the any two values.

6. 5. The shape detection system according to claim 4, A shape detection system that calculates the tilt direction based on the ratio of the difference between any two of the RGB values to the difference between any two values different from the any two of the RGB values.

Images