Tail clearance measurement system and tail clearance measurement method

Abstract

To provide a tail clearance measurement system and a tail clearance measurement method capable of measuring a tail clearance with high accuracy.SOLUTION: A tail clearance measurement system 1 for measuring a tail clearance between a skin plate 3 and a segment 4 of a shield excavator 2 comprises information acquisition means 10 having imaging means and depth information acquisition means, and analysis means 20. The analysis means 20 comprises: segment reference point range specifying means 22 for specifying a range in which the reference point of the segment 4 is included in an image based on the parallax amount and the three-dimensional shape of the joint surface 4b of the segment 4; and segment reference point specifying means 23 for performing image processing on the range of the image and specifying the reference point of the segment 4 on the image.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a tail clearance measurement system and a tail clearance measurement method. [Background technology]

[0002] One of the construction management items for shield tunneling is "tail clearance." Tail clearance is the distance between the outer surface of the segments assembled inside the tail of the shield tunneling machine and the inner surface of the steel shell (skin plate) of the shield tunneling machine. Conventionally, when shield tunneling or segment assembly is completed, workers move to a specified location and measure the tail clearance using a measuring tool (e.g., a convex or vernier caliper). In addition, there are techniques for measuring tail clearance using, for example, a camera or a line laser measuring instrument (see, for example, Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2002-023632 A [Patent Document 2] JP 2009-025252 A Summary of the Invention [Problem to be solved by the invention]

[0004] When workers use measuring tools to measure tail clearance, they need to move to the measurement location, which reduces the efficiency of the work. In addition, when the measurement location is high up, the work must be done carefully, which takes more time than when the measurement location is low. Also, when measuring tail clearance using a camera or the like, there are cases where measurement is not possible depending on the shooting conditions. For example, as shown in Figure 16(b), when a wet skin plate reflects light and is recognized as the same color as the segments, or as shown in Figure 16(c), when the tail grease is recognized as the same color as the segments, measurement with a camera is not suitable. Note that Figure 16(a) shows a state where tail clearance can be measured. From this perspective, the present invention provides a tail clearance measurement system and a tail clearance measurement method that are capable of measuring tail clearance with high accuracy. [Means for solving the problem]

[0005] The tail clearance measurement system according to the present invention is a system for measuring the tail clearance between a skin plate and a segment of a shield machine. The tail clearance measurement system includes an imaging means, a depth information acquisition means, and an analysis means. The imaging means captures an image including the skin plate and the segment. The depth information acquisition means acquires depth information relating to a distance to a joint surface of the segment. The analysis means identifies positions of the segment and the skin plate on the image using the depth information, and calculates a tail clearance based on a positional relationship between the skin plate and the segment on the image. The analysis means includes a segment reference point range identification means, a segment reference point identification means, and a tail clearance calculation means. The segment reference point range identifying means identifies a range in the image including the reference point of the segment based on the amount of parallax between the imaging means and the depth information acquisition means and the three-dimensional shape of the joint surface recognizable based on the depth information. The segment reference point identifying means performs image processing on the range of the image and identifies the reference point of the segment on the image. The plate reference point identifying means identifies the reference point of the skin plate on the image based on the depth information and the angle of view of the imaging means. The tail clearance calculation means calculates the tail clearance based on the number of pixels between the reference point of the segment and the reference point of the skin plate on the image.

[0006] In the tail clearance measurement system according to the present invention, a range including the reference point of the segment on the image is specified based on information acquired by the depth information acquisition means, and the reference point on the image is specified by performing image processing on the range. Therefore, the image processing can be performed without being affected by disturbances outside the specified range. As a result, erroneous recognition is less likely to occur than when the reference point is directly recognized from the image.

[0007] The apparatus may further include a storage means for storing design data relating to the joint surface. In this case, the segment reference point range specifying means superimposes a cross-sectional shape at an arbitrary cross section of the three-dimensional shape and a cross-sectional shape of the joint surface in the design data by adjusting a scale and a position, and determines a reference point of the joint surface in the design data as a reference point of the joint surface in the three-dimensional shape. Then, the determined reference point of the three-dimensional shape may be corrected by the parallax amount to specify a range on the image that includes the reference point of the segment. The segment reference point range specifying means may determine completion of the superposition by a regression analysis method. The depth information acquiring means may be a depth camera using a structured illumination method or a time-of-flight method. The segment reference point specifying means may specify the reference point of the segment on the image by a spatial high frequency detection filter or a Hough transform.

[0008] The imaging means may be configured to have at least a first imaging means for close-range photography set to an angle of view capable of capturing an image of the area around the gap between the skin plate and the segment when the shield jack is in a minimum stroke state, and a second imaging means for long-range photography set to an angle of view capable of capturing an image of the area around the gap when the shield jack is in a maximum stroke state. In this way, since the autofocus function is not used, errors caused by the mechanism for implementing the function can be eliminated, and therefore the tail clearance can be measured with high accuracy.

[0009] The tail clearance measurement method according to the present invention is a method for measuring the tail clearance between a skin plate and a segment of a shield machine. This tail clearance measurement method includes an information acquisition step and an analysis step. In the information acquisition step, an image including the skin plate and the segment and depth information regarding the distance to the joint surface of the segment are acquired by the imaging means and the depth information acquisition means. In the analysis step, the positions of the segment and the skin plate on the image are identified using the depth information, and a tail clearance is calculated based on the positional relationship between the skin plate and the segment on the image. The analysis step includes a segment reference point range identification step, a segment reference point identification step, a plate reference point identification step, and a tail clearance calculation step. In a segment reference point range identifying process, a range in the image containing the reference point of the segment is identified based on the amount of parallax between the imaging means and the depth information acquisition means and the three-dimensional shape of the joint surface recognizable based on the depth information. In a segment reference point identifying process, image processing is performed on the range of the image to identify the reference point of the segment on the image. In a plate reference point identifying process, a reference point of the skin plate is identified on the image based on the depth information and the angle of view of the imaging means. In a tail clearance calculation process, a tail clearance is calculated based on the number of pixels between the reference point of the segment and the reference point of the skin plate on the image.

[0010] In the tail clearance measurement method according to the present invention, a range including the reference point of the segment on the image is specified based on information acquired by the depth information acquisition means, and the reference point on the image is specified by performing image processing on the range. Therefore, the image processing can be performed without being affected by disturbances outside the specified range. As a result, erroneous recognition is less likely to occur than when the reference point is directly recognized from the image. Effect of the Invention

[0011] According to the present invention, the tail clearance can be measured with high accuracy. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is a configuration diagram of a tail clearance measurement system according to an embodiment of the present invention. [Diagram 2] FIG. 4 is a schematic diagram showing the positional relationship between an information acquisition means, a skin plate, and a segment. [Diagram 3] 4 is an example of an image captured by an imaging unit. [Figure 4] FIG. 2 is a diagram for explaining the imaging means, where (a) is an image captured by a first imaging means, and (b) is an image captured by a second imaging means. [Diagram 5] A figure for explaining the processing of the segment reference point range identification means, where (a) is an example of a depth image, and (b) is a cross-sectional image of the depth image, which is a cross-sectional image corresponding to VB-VB in Figure 5(a). [Figure 6] 11A and 11B are diagrams for explaining the processing of the segment reference point range identification means, where (a) shows the process of superimposing the cross-sectional shape of the design data on the cross-sectional shape of the measured segment, and (b) shows the state in which the cross-sectional shape of the design data is superimposed on the cross-sectional shape of the measured segment. [Figure 7] 5A and 5B are diagrams for explaining the processing of a segment reference point range specifying means, where (a) is an example of a depth image, and (b) and (c) are example images. [Figure 8]1A and 1B are diagrams for explaining edge detection processing, in which FIG. 1A is an image of processing using a vertical spatial high frequency detection filter, and FIG. 1B is an image of Hough transform processing. [Figure 9] This figure explains the processing of the plate reference point identification means, where (a) shows the case where the distance to the reference point is "4 m", (b) shows the case where the distance to the reference point is "3 m", and (c) shows the position of the reference point in the image according to the distance. [Figure 10] FIG. 11 is a diagram for explaining a method for calculating the position of a reference point on an image, where (a) is a schematic diagram showing the relationship between the position and angle of view of the imaging means, the position of the skin plate, and the position of the reference point, and (b) is a diagram showing the relationship between distance and the reference point in an image. [Figure 11] 10A and 10B are diagrams for explaining a method for calculating the position of a reference point on an image, where (a) is an example of the angle of view of a lens of an imaging means, and (b) is a table showing the relationship between distance and the vertical coordinate of a reference point. [Figure 12] FIG. 4A is a diagram showing the relationship between the angle of view of the imaging means and the height of the field of view, and FIG. 4B is a diagram for explaining the processing of the tail clearance calculation means. [Figure 13] 1 is an example of a flowchart illustrating steps of a tail clearance measurement method. [Figure 14] 1A and 1B are diagrams for explaining the contents of a simulation conducted to verify the effect of a tail clearance measurement system according to an embodiment of the present invention, in which (a) is a schematic diagram of the tail clearance measurement system, (b) is the lens angle of view, (c) is a table showing the relationship between distance and length per pixel, and (d) is a table showing the relationship between distance and tail clearance. [Figure 15] 1A and 1B are diagrams for explaining the effects of the tail clearance measurement system according to an embodiment of the present invention, in which (a) shows a case in which a reference point is erroneously recognized using a conventional RGB camera, and (b) shows a case in which a reference point is set at a position where there is no actual object. [Figure 16]1A and 1B are diagrams for explaining the problem, in which (a) shows a state in which the tail clearance can be measured, and (b) and (c) show states in which the tail clearance cannot be measured. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, the embodiment of the present invention will be described in detail with reference to the drawings. Each drawing is merely a schematic illustration to allow a sufficient understanding of the present invention. Therefore, the present invention is not limited to the illustrated examples. In each drawing, common or similar components are denoted by the same reference numerals, and duplicated explanations thereof will be omitted.

[0014] <Configuration of tail clearance measurement system according to embodiment> The configuration of a tail clearance measurement system 1 according to an embodiment will be described with reference to Figures 1 and 2. Figure 1 is a configuration diagram of the tail clearance measurement system 1. Figure 2 is a schematic diagram showing the positional relationship between an information acquisition means 10, a skin plate 3, and a segment 4. The tail clearance measurement system 1 shown in Fig. 1 is a system capable of automatically measuring tail clearance during shield tunneling. The shield tunneling machine 2 rotates the cutter 5 attached to the front end and presses the shield jack 6 against the joint surface (front end surface) 4b of the segment 4 attached to the rear, obtaining a propulsive force to excavate the ground. A plurality of shield jacks 6 (only the top and bottom of the shield tunneling machine 2 are shown in Fig. 1) are arranged at a predetermined pitch in the circumferential direction.

[0015] The skin plate 3 shown in Fig. 1 is a cylindrical member. At the rear of the skin plate 3, a segment ring is formed by connecting a plurality of segments 4 in the circumferential direction. A gap is formed between the inner peripheral surface 3a of the skin plate 3 and the outer peripheral surface 4a of the segments 4. A tail seal 8 is provided on the inside of the rear end of the skin plate 3. The tail seal 8 prevents groundwater, soil, backfill material, etc. from flowing into the shield excavator 2 through the gap between the skin plate 3 and the segments 4. The tail seal 8 is made of, for example, a wire brush, and is filled with grease.

[0016] The tail clearance is the distance between the outer peripheral surface 4a of the segment 4 in the tail of the shield excavator 2 and the inner peripheral surface 3a of the skin plate 3 of the shield excavator 2, and is one of the construction management items of the shield construction. For example, the distance between the reference point 4c set on the segment 4 shown in FIG. 2 and the reference point 3c set on the skin plate 3 is measured. The reference point 4c on the segment 4 side is set at the boundary between the outer peripheral surface 4a and the joint surface (front end surface) 4b, and the reference point 3c on the skin plate 3 is a position radially outward from the reference point 4c. In the segment 4 in this embodiment, a step shape is formed at the upper and lower edges of the joint surface 4b, and the reference point 4c is set at the boundary between the outer peripheral surface 4a and a vertical surface continuing from the outer peripheral surface 4a.

[0017] As shown in FIG. 1, the tail clearance measurement system 1 includes one or more information acquisition means 10 and an analysis means 20. The information acquisition means 10 and the analysis means 20 are installed inside the shield tunneling machine 2. The information acquisition means 10 and the analysis means 20 are connected so that data communication is possible. The data communication method is not particularly limited and may be either wired or wireless. The analysis means 20 may also be installed outside the shield tunneling machine 2 (for example, in an office outside the tunnel). The information acquisition means 10 may also be configured to have the functions of the analysis means 20 (i.e., the information acquisition means 10 and the analysis means 20 may be configured as a single device).

[0018] (Information acquisition means) The information acquisition means 10 shown in FIG. 1 acquires information necessary for calculating the tail clearance. The information acquisition means 10 in this embodiment is installed at the mounting position of the shield jack 6 (for example, on the ring girder). As shown in FIG. 2, the information acquisition means 10 has a depth information acquisition means 11 and at least one or more imaging means 12 (two imaging means, a first imaging means 12A and a second imaging means 12B, in this embodiment). In this embodiment, the information acquisition means 10 is assumed to be a single device in which the depth information acquisition means 11 and the imaging means 12 are housed in one housing, but these components may be housed in separate housings to constitute multiple devices. The information acquisition means 10 is disposed at a position offset in the circumferential direction with respect to the shield jack 6, and acquires images of the segments 4 and depth information of the segments 4 from between the spreaders 6a.

[0019] The depth information acquisition means 11 can acquire depth information related to the depth, which is the distance from the reference point of the depth information acquisition means 11 to the object. The depth information acquisition means 11 is, for example, a depth camera, and in this embodiment, a depth camera of the "structured illumination method" or the "Time of Flight (ToF) method" is assumed as the depth information acquisition means 11. In this case, the depth is the distance from the lens to the object. The pixel value of the depth image (depth image) acquired by this depth camera indicates the distance to the object. The depth information acquisition means 11 is installed so as to be able to acquire information on the periphery of the gap 7 between the skin plate 3 and the segment 4 shown in FIG. 2. The depth information acquisition means 11 outputs the acquired depth information to the analysis means 20. It is also possible to use a 3D scanner or a laser range finder as the depth information acquisition means 11. The depth information acquisition means 11 may also acquire depth information from an excavation management system 40 (see FIG. 1). The stroke value of the shield jack 6 is measured by another means (for example, a stroke meter built into the jack or a laser distance meter), and the measured stroke value is provided to the excavation management system 40. The excavation management system 40 obtains the distance (depth information) to the segment 4 from the provided stroke value, and transmits the depth information to the depth information acquisition means 11.

[0020] The imaging means 12 is capable of capturing an image of an object, and is, for example, an RGB camera. The imaging means 12 is installed so as to capture the same object as the depth information acquisition means 11, and in this embodiment, is installed so as to capture the gap 7 between the skin plate 3 and the segment 4 and its surroundings shown in Fig. 2. The depth image acquired by the depth information acquisition means 11 and the RGB image acquired by the imaging means 12 are arranged so that there is no misalignment in the positions on the images (for example, correction is performed to align the positions of the two images), and since the resolution of the depth information acquisition means 11 is insufficient for the accuracy of clearance measurement, the measurement accuracy is improved by combining with a high-resolution imaging means 12.

[0021] The first imaging means 12A and the second imaging means 12B shown in Fig. 2 have fixed focal points, and are set to different focal lengths L (see Fig. 4). Fig. 4(a) is an image captured by the first imaging means 12A, and (b) is an image captured by the second imaging means 12B. For example, as shown in Fig. 4(a), the first imaging means 12A has a short focal length LA, and the angle of view is set to allow the shield jack 6 (see Fig. 1) to properly capture an image of the periphery of the gap 7 when the shield jack 6 (see Fig. 1) is at its minimum stroke (i.e., the setting is for close-up photography). On the other hand, the second imaging means 12B has a long focal length LB, and the angle of view is set to allow the shield jack 6 (see Fig. 1) to properly capture an image of the periphery of the gap 7 when the shield jack 6 (see Fig. 1) is at its maximum stroke (i.e., the setting is for long-distance photography). The imaging means 12 outputs the created image data to the analysis means 20.

[0022] (Analysis means) The analysis means 20 shown in FIG. 1 is, for example, a computer including a central processing unit (CPU), a read only memory (ROM), a random access memory (RAM), and the like. The analysis means 20 calculates the tail clearance based on the positional relationship between the skin plate 3 and the segment 4 on the image captured by the imaging means 12. Specifically, the tail clearance is calculated based on the distance between the reference point 4c of the segment 4 and the reference point 3c of the skin plate 3, which are shown in the image F shown in FIG. 3. FIG. 3 is an example of the image F captured by the imaging means 12. The positions of the reference points 3c, 4c on the image F are identified by analysis using information acquired by the depth information acquisition means 11, and the like. 1, the analysis means 20 includes a storage means 21, a segment reference point range specifying means 22, a segment reference point specifying means 23, a plate reference point specifying means 24, and a tail clearance calculation means 25. The functions of the analysis means 20 are realized, for example, by a CPU executing a program.

[0023] Information required for calculating the tail clearance is stored in the storage means 21. For example, the storage means 21 stores design data related to the joint surface 4b of the segment 4. The design data related to the joint surface 4b includes, for example, information related to the reference point 4c. The storage means 21 also stores information regarding the amount of parallax between the depth information acquisition means 11 and the imaging means 12. The depth information acquisition means 11 and the imaging means 12 are arranged, for example, on a plane perpendicular to the front-rear direction (on a plane consisting of up-down and left-right directions), and the optical axis is decentered. In addition, the depth information acquisition means 11 and the imaging means 12 use different lenses, so there are differences in the angle of view and lens distortion. Information for correcting the difference in images caused by these factors, the "amount of parallax," is stored. The amount of parallax becomes known by performing calibration in which specific subjects are photographed simultaneously and the relationship is expressed as a function. The storage means 21 also stores information regarding the positional relationship between the depth information acquisition means 11 and the imaging means 12 and the skin plate 3. The depth information acquisition means 11 and the imaging means 12 are installed at fixed points and their positional relationship with the skin plate 3 is constant. The information is, for example, the distance from the skin plate 3 to the depth information acquisition means 11 and the imaging means 12.

[0024] The calculation process of the tail clearance in the analysis means 20 will be described with reference to Figures 5 to 12 (and Figures 1 to 4 as appropriate). Here, attention will be focused on the segment 4 (see Figure 1) that is located at the bottom, and the description will be given assuming that the reference point 4c of the segment 4 is the position of the lowest point (lower end position). The segment reference point range specifying means 22 shown in FIG. 1 specifies an approximate range of the reference point 4c of the segment 4 on the image F (a range in which the reference point 4c is likely to be included) based on the three-dimensional shape of the joint surface 4b that can be recognized based on the depth information acquired by the depth information acquiring means 11. Specifically, the depth image G1 shown in FIG. 5(a) is cut at an arbitrary cross section, and a cross-sectional image 104 (cross-sectional shape) of the three-dimensional shape of the joint surface 4b is created as shown in FIG. 5(b). FIG. 5 is a diagram for explaining the processing of the segment reference point range specifying means 22, where (a) is an example of a depth image, and (b) is a cross-sectional image of the depth image, which corresponds to VB-VB in FIG. 5(a). The cross section cutting the depth image G1 is preferably perpendicular to the horizontal plane (a plane formed by the front, back, left and right), and in that case, a cross-sectional image 104 (cross-sectional shape) of the joint surface 4b viewed from the side is created (see FIG. 5(b)). Since the depth information acquisition means 11 cannot capture an image of the hidden surface of the segment 4, the cross-sectional image 104 shown in Fig. 5(b) shows only the shape of the vicinity of the joint surface 4b of the segment 4. The reference numeral 103 shown in Fig. 5(b) is a cross-sectional image (cross-sectional shape) of the three-dimensional shape of the inner peripheral surface 3a of the skin plate 3.

[0025] Next, the segment reference point range specifying means 22 sets the cross-sectional image 204 (cross-sectional shape) of the joint surface 4b stored in the storage means 21 to the depth image G2 (see FIG. 6(a)). The cross-sectional image 204 shown in FIG. 6(a) is the cross-sectional shape of the segment 4 based on the design data, for example, obtained by cutting a three-dimensional model of the segment 4 at an arbitrary cross section. Information on the reference point 4c is associated with the cross-sectional image 204. At this time, based on the distance to the joint surface 4b (depth information) and the angle of view of the depth information acquiring means 11, the size of the cross-sectional image 104 or the cross-sectional image 204 is changed to the size of the other (scale conversion) to make the sizes of the cross-sectional image 104 and the cross-sectional image 204 the same.

[0026] Then, as indicated by the symbol α in FIG. 6(a), the cross-sectional image 204 based on the design data is moved and superimposed on the cross-sectional image 104 based on the actual image (see FIG. 6(b)). That is, the segment reference point range specifying means 22 adjusts the scale and position to superimpose the cross-sectional image 204 and the cross-sectional image 104. In FIG. 6, the cross-sectional image 204 based on the design data is moved in the vertical direction along the extension line of the joint surface 4b of the cross-sectional image 104 acquired by photographing. The criterion for determining that the two cross-sectional shapes (the cross-sectional image 104 acquired by photographing and the cross-sectional image 204 based on the design data) match can be, for example, a regression analysis method such as the least squares method. That is, the completion of the superimposition of the two cross-sectional shapes is determined by the regression analysis method. Note that the match may be determined by using a method other than the regression analysis method. When the two cross-sectional shapes (the cross-sectional image 104 acquired by photographing and the cross-sectional image 204 based on the design data) are superimposed, the position of the reference point 4c in the cross-sectional image 104 is specified. In other words, by superimposing the two cross-sectional shapes, it is possible to reflect the position of reference point 4c associated with cross-sectional image 204 in cross-sectional image 104, thereby identifying the position of reference point 4c in depth image G1 (see Figure 5(a)).

[0027] Since the optical axes of the depth information acquisition means 11 and the imaging means 12 are decentered, the position of the segment 4 is shifted between the depth image G1 (see FIG. 7(a)) captured by the depth information acquisition means 11 and the image F (see FIG. 7(b)) captured by the imaging means 12. In addition, since the angles of view are different, the segment 4 appears at different sizes in the depth image G1 (see FIG. 7(a)) and the image F (see FIG. 7(b)). Therefore, the position of the reference point 4c identified in the depth image G1 cannot be reflected directly in the image F. Therefore, the segment reference point range identification means 22 considers the differences between the images (parallax amount) based on differences in the angles of view, lens distortion, optical axis, etc., and identifies the approximate range including the reference point 4c of the segment 4 on the image F by correcting according to the parallax amount. In Fig. 7(b), the position corresponding to the reference point 4c in the depth image G1 (see Fig. 7(a)) is indicated by the symbol 4ca, and the range in which the position 4ca is corrected using the amount of parallax is indicated by the symbol 4cb. The amount of parallax can be grasped, for example, by simultaneously photographing a specific subject using the depth information acquisition means 11 and the imaging means 12, and performing calibration to convert the relationship into a function. It is preferable to grasp the amount of parallax in advance and register it in the storage means 21 in advance.

[0028] Next, the segment reference point identification means 23 (see FIG. 1) detects the reference point 4c in the image F (see FIG. 7(c)) by performing image processing on the range 4cb identified in the image F (see FIG. 7(b)). The image processing for detecting the reference point 4c in the image F is not particularly limited, and existing technology can be used. For example, the edge (boundary) between the lower end of the segment 4 and the space is detected, and the reference point 4c is identified from the detected edge. For example, a vertical spatial high frequency detection filter or a Hough transform process may be used to detect the edge (boundary). FIG. 8(a) is an image of processing using a vertical spatial high frequency detection filter, and FIG. 8(b) is an image of Hough transform processing.

[0029] In the filter process shown in FIG. 8(a), a portion where the brightness of the image changes suddenly is detected as an edge. For example, assuming an xy coordinate system with the origin of the y coordinate at the top end of image F, the bottom end of segment 4 is detected as edge 4cc using a vertical spatial high frequency detection filter 301. For example, edge: E xy , pixel value P xy In this case, "E xy =|P xy+1 -P xy-1 ", E xy For a set Y of y where is greater than a certain threshold, maxY is set as the reference point 4c of segment 4. 8(b), since it is possible to extract straight line components in an image, a straight line 4cd constituting the bottom end of the segment 4 is detected. The y coordinate of an arbitrary position on the detected straight line 4cd is set as a reference point 4c of the segment 4. By the process described above, it is possible to identify the reference point 4c in the image F with high accuracy (for example, a resolution of 0.5 mm or less).

[0030] Next, the plate reference point identification means 24 (see FIG. 1) determines the position of the reference point 3c of the skin plate 3 in the image F. For example, assume that the angle of view of the imaging means 12 is constant and the direction of the optical axis is fixed. Comparing the case where the distance from the imaging means 12 to the reference point 3c of the skin plate 3 is "4m" shown in FIG. 9(a) with the case where the distance from the imaging means 12 to the reference point 3c of the skin plate 3 is "3m" shown in FIG. 9(b), the reference point 3c is projected closer to the center of the image F when the distance shown in FIG. 9(a) is longer. In other words, as shown in FIG. 9(c), the reference point 3c of the skin plate 3 is located closer to the center of the image F as the distance to the segment 4 becomes longer, and the position of the reference point 3c moves to the outside of the image F as the distance to the segment 4 becomes closer. In this way, if the distance to the segment 4 is known, the position of the reference point 3c of the skin plate 3 in the image F can be identified. Although the angle of view of the imaging means 12 varies, the correct position can be calculated by calibrating the difference between the theoretical position of the reference point 3c and the actual position of the reference point 3c at each distance during manufacturing. Also, it is advisable to correct the variation in the optical axis by adjusting the imaging means 12 when it is installed.

[0031] With reference to FIG. 10(a), calculation of the position of reference point 3c located at distance D in image F will be described. Here, it is assumed that the lens angle of view θ of the imaging means 12 is known and the optical axis is parallel to the skin plate 3. Also, as shown in FIG. 10(b), an xy coordinate system with the upper left corner of image F as the origin is set in the image F. FIG. 10 is a diagram for explaining a method of calculating the position of reference point 3c on image F, where (a) is a schematic diagram showing the relationship between the position and angle of view of the imaging means 12, the position of the skin plate 3, and the position of reference point 3c, and (b) is a diagram showing the relationship between distance D and reference point 3c in image F.

[0032] At a distance D, the length dl of one pixel is calculated by "dl = D x tan(θ / 2) x 2 / number of vertical pixels". Moreover, the number of pixels dp from the center (optical axis) of the imaging means 12 to the reference point 3c is calculated by "dp=H / dl". Also, the vertical coordinate v_fl of the reference point 3c is calculated by "v_fl = number of vertical pixels / 2 + dp = number of vertical pixels / 2 + H / (D x tan(θ / 2) x 2 / number of vertical pixels)". Here, the distance D is known from the measurement results (depth information) by the depth information acquisition means 11, and the distance H is known by installing the imaging means 12 at a fixed point.

[0033] Assume that the installation height of the imaging means 12 is "H=0.3 m", the number of horizontal pixels of the image sensor is "3840", and the number of vertical pixels is "2880". As shown in FIG. 11(a), the focal length of the lens of the first imaging means 12A for close-up photography is "6 mm", and the vertical angle of view is "42.9510°". The focal length of the lens of the second imaging means 12B for long-distance photography is "12 mm", and the vertical angle of view is "22.2569°". Note that only the vertical direction is considered here, and the horizontal angle of view is not mentioned. The vertical coordinate v_fl on the image F of the reference point 3c of the skin plate 3 calculated under these conditions is as shown in the table shown in Fig. 11(b). In the table shown in Fig. 11(b), the vertical coordinate v_fl has a different value depending on the distance D. The smaller the distance D, the larger the vertical coordinate v_fl, and the larger the distance D, the smaller the vertical coordinate v_fl. In this manner, the plate reference point identifying means 24 calculates the position of the reference point 3c in the image F.

[0034] 12(a), when the angle of view of the imaging means 12 and the distance from the imaging means 12 to the joint surface 4b of the segment 4 are known, the height of the field of view of the captured range is determined, and the length per pixel can be determined by dividing the field of view by the number of vertical pixels (e.g., 3000 pixels). If the angle of view varies, it is advisable to perform calibration by photographing a specific subject. 12(b), the tail clearance can be calculated by multiplying the number of pixels between the reference point 4c of the segment 4 and the reference point 3c of the skin plate 3 by the length per pixel. In this manner, the tail clearance calculation means 25 calculates the tail clearance from the image F.

[0035] <Tail clearance measurement method according to embodiment> A method for measuring tail clearance using the tail clearance measurement system 1 will be described with reference to Fig. 13 (and with reference to Figs. 1 to 12 as appropriate). Fig. 13 is an example of a flowchart showing steps of the tail clearance measurement method. First, the depth information acquisition means 11 (one example is a depth camera) captures an image of the gap 7 (see FIG. 2) between the skin plate 3 and the segment 4 and its surroundings, and the segment reference point range identification means 22 detects the reference point 4c of the segment 4 using the depth image (step S11). As shown in FIG. 6, in detecting the reference point 4c, a cross-sectional image 204 based on the design data is superimposed on a cross-sectional image 104 of the joint surface (front end surface) 4b acquired by actually photographing it. Then, the position of the reference point 4c set in the cross-sectional image 204 is reflected in the cross-sectional image 104. In addition, the distance (depth information) to the reflected reference point 4c is read from the depth image.

[0036] Next, the imaging means 12 to be used is selected according to the distance to the reference point 4c of the segment 4 (step S12). As shown in FIG. 4(a), when the distance to the reference point 4c is closer than two preset lens switching positions, the first imaging means 12A for close-up photography is selected. Also, as shown in FIG. 4(b), when the distance to the reference point 4c is farther than the switching position, the second imaging means 12B for long-distance photography is selected. Then, the selected imaging means 12 (either the first imaging means 12A or the second imaging means 12B) is used to image the periphery of the gap 7 (see FIG. 2) between the skin plate 3 and the segment 4. Note that in step S11, the first imaging means 12A and the second imaging means 12B may simultaneously capture images with the depth information acquisition means 11, and the image of one of the imaging means 12 selected in step S12 may be adopted (in that case, the image of the other imaging means 12 that was not selected is discarded).

[0037] Next, the distance between the reference point 4c of the segment 4 and the reference point 3c of the skin plate 3 is calculated by performing image analysis on the imaging data acquired by the imaging means 12 (step S13). For example, as shown in FIG. 7(b), an approximate range 4cb of the reference point 4c in the image F is specified based on the amount of parallax, and the position of the reference point 4c in the image F is detected by performing image processing on the range 4cb (see FIG. 7(c)). An example of the image processing for detecting the reference point 4c is processing such as a vertical spatial high frequency detection filter or Hough transformation. In addition, since the distance to the reference point 4c of the segment 4 is known by the processing of step S11, the position of the reference point 3c of the skin plate 3 in the image F is specified based on the distance to the reference point 4c, as shown in FIG. 9. Then, as shown in FIG. 12(b), the tail clearance is calculated by multiplying the number of pixels between the reference point 4c of the segment 4 and the reference point 3c of the skin plate 3 by the length per pixel.

[0038] The calculation result (tail clearance) in step S13 may be output to the excavation management system 40 (see FIG. 1) and the calculation result may be visualized (step S14). There are no limitations on the visualization method, and it is preferable to display the tail clearance so that an operator can confirm it at a glance, for example. The process of steps S11 to S14 is repeatedly executed during excavation.

[0039] As described above, in the tail clearance measurement system 1 according to this embodiment, the range 4cb (see FIG. 7) including the reference point 4c of the segment 4 on the image F is identified based on the information acquired by the depth information acquisition means 11, and image processing is performed on the range 4cb to identify the reference point 4c on the image F. Therefore, the image processing can be performed without being affected by disturbances that fall outside the identified range 4cb. As a result, erroneous recognition is less likely to occur than when the reference point 4c is directly recognized from the image F. Furthermore, the tail clearance measurement system 1 according to this embodiment does not use an autofocus function, so errors caused by a mechanism for implementing this function can be eliminated, making it possible to measure the tail clearance with high accuracy.

[0040] A simulation was carried out to verify the effects of the tail clearance measurement system 1 according to this embodiment, and the results are shown in Fig. 14. Fig. 14 is a diagram for explaining the contents of the simulation carried out to verify the effects, in which (a) is a schematic diagram of the tail clearance measurement system 1, (b) is the angle of view of the lens, (c) is a table showing the relationship between distance and length per pixel, and (d) is a table showing the relationship between distance and tail clearance.

[0041] In the simulation, the effect on the clearance (tail clearance error) when an error of "±2%" occurs in the measurement distance of the depth information acquisition means 11 (one example is a depth camera) was calculated. The conditions in the simulation are as shown in Figs. 14(a) to 14(c). For example, as shown in Fig. 14(b), the focal length of the lens of the first imaging means 12A for close-up photography is "6 mm", the horizontal angle of view is "55.2911°", and the vertical angle of view is "42.9510°". In addition, the focal length of the lens of the second imaging means 12B for long-distance photography is "12 mm", the horizontal angle of view is "29.3541°", and the vertical angle of view is "22.2569°". In addition, as shown in Fig. 14(c), the number of horizontal pixels of the image sensor is "3840", and the number of vertical pixels is "2880". As shown in Figure 14(d), the simulation results for both short and long distances were within ±1 mm for the tail clearance distance of 40 mm, confirming that the system is fully usable at actual construction sites.

[0042] Furthermore, according to the tail clearance measurement system 1 of this embodiment, even if the joint surface 4b of the segment 4 is not flat but has unevenness, the reference point 4c can be detected after correctly recognizing the unevenness with the depth information acquisition means 11 (one example is a depth camera). Furthermore, when using only a conventional RGB camera, as shown in FIG. 15(a), there is a possibility that the reference point 4c will be erroneously recognized if the joint surface 4b is chamfered, in which case the tail clearance cannot be measured accurately. Moreover, according to the tail clearance measurement system 1 of this embodiment, the joint surface 4b can be detected three-dimensionally by the depth information acquisition means 11 (one example is a depth camera). Therefore, as shown in Fig. 15(b), it is also possible to set the reference point 4c at a position where there is no actual object, and it is possible to deal with various situations. [Explanation of symbols]

[0043] 1 Tail Clearance Measurement System 2. Shield tunneling machine 3. Skin Plate 3a Inner surface 3c Reference point 4 Segments 4a Outer surface 4b Joint surface 4c Reference point 5. Cutter 6 Shield Jack 10 Information acquisition means 11 Depth information acquisition means 12 Imaging means 12A First Imaging Means 12B Second imaging means 20 Analysis methods 21 Memory means 22 Segment reference point range determination means 23 Segment Reference Point Identification Means 24 Plate reference point identification means 25 Tail clearance calculation method

Claims

1. A tail clearance measurement system for measuring a tail clearance between a skin plate and a segment of a shield excavator, comprising: An imaging means for capturing an image including the skin plate and the segment; A depth information acquisition means for acquiring depth information relating to a distance to a joint surface of the segment; and an analysis means for determining the tail clearance based on a positional relationship between the skin plate and the segment on the image, The analysis means identifies the positions of the segments and the skin plate on the image using the depth information. A tail clearance measurement system comprising:

2. The analysis means a segment reference point range specifying means for specifying a range in the image that includes a reference point of the segment based on a parallax amount between the imaging means and the depth information acquiring means and a three-dimensional shape of the joint surface that can be recognized based on the depth information; a segment reference point specifying means for performing image processing on the range of the image and specifying a reference point of the segment on the image; a plate reference point specifying means for specifying a reference point of the skin plate on the image based on the depth information and an angle of view of the imaging means; and a tail clearance calculation means for calculating a tail clearance based on the number of pixels between a reference point of the segment on the image and a reference point of the skin plate.

2. The tail clearance measurement system according to claim 1 .

3. Further comprising a storage means for storing design data relating to the joint surface, The segment reference point range specifying means A cross-sectional shape at an arbitrary cross section of the three-dimensional shape and a cross-sectional shape of the joint surface in the design data are superimposed by adjusting the scale and position, and a reference point of the joint surface in the design data is determined as a reference point of the joint surface in the three-dimensional shape; correcting the determined reference point of the three-dimensional shape using the amount of parallax, thereby specifying a range on the image that includes the reference point of the segment; 3. The tail clearance measurement system according to claim 2, wherein the tail clearance measurement system is a tail clearance measurement system.

4. 4. The tail clearance measurement system according to claim 3, wherein the segment reference point range specifying means determines completion of the overlapping by a regression analysis method.

5. 2. The tail clearance measurement system according to claim 1, wherein the depth information acquisition means is a depth camera using a structured illumination method or a time-of-flight method.

6. the segment reference point specifying means specifies the reference point of the segment on the image by a spatial high frequency detection filter or a Hough transform; 3. The tail clearance measurement system according to claim 2, wherein the tail clearance measurement system is a tail clearance measurement system.

7. The imaging means is a first imaging means for close-range photography set to an angle of view capable of photographing the periphery of the gap between the skin plate and the segment when the shield jack is in a minimum stroke state; and a second imaging means for long-distance photography, the second imaging means being set to an angle of view capable of photographing the periphery of the gap when the shield jack is in a maximum stroke state.

2. The tail clearance measurement system according to claim 1 .

8. A tail clearance measurement method for measuring a tail clearance between a skin plate and a segment of a shield machine, comprising the steps of: an information acquiring step of acquiring an image including the skin plate and the segment and depth information related to a distance to a joint surface of the segment by an imaging means and a depth information acquiring means; and an analysis step of determining the tail clearance based on a positional relationship between the skin plate and the segment on the image, In the analyzing step, the depth information is used to identify the positions of the segment and the skin plate on the image. A tail clearance measurement method comprising:

9. The analyzing step includes: a segment reference point range specifying step of specifying a range in the image that includes a reference point of the segment based on a parallax amount between the imaging means and the depth information acquisition means and a three-dimensional shape of the joint surface that can be recognized based on the depth information; a segment reference point specifying step of performing image processing on the range of the image and specifying a reference point of the segment on the image; a plate reference point specifying step of specifying a reference point of the skin plate on the image based on the depth information and the angle of view of the imaging means; A tail clearance calculation step of calculating a tail clearance based on the number of pixels between a reference point of the segment and a reference point of the skin plate on the image.

9. The tail clearance measuring method according to claim 8,

Images