JP2013015452A - Reinforcement range extraction device, reinforcement range extraction method, and reinforcement range extraction program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simply and accurately acquire information on reinforcement arrangement.SOLUTION: In a reinforcement information acquisition system 1, a portable terminal 4 detects markers MK1 and MK2 from the whole image corresponding to a part of a background bar B (S901). Next, an image existing between the marker MK1 and the marker MK2 is cut out (S902). In the cut image, by detecting locations where the brightness of pixel is discontinued in a direction perpendicular to the axis direction of a reinforcement, edges serving as area boundary lines are extracted (S903). Then, from the extracted edges, unnecessary edges are removed (S904). Next, frequency distribution for brightness distribution of the image is created, and based on the frequency distribution, representative brightness values in three ranges of reinforcement, shadow and background are calculated, and using the representative brightness values, the range encircled by respective edges is determined as any of the three (S905). Furthermore, when among the determined ranges, two adjacent ranges sandwiching an edge are the same, the two ranges are integrated as one range (S906).

Description

  The present invention relates to an apparatus, method, and program for extracting a reinforcing bar region from a captured image including a reinforcing bar, a shadow of the reinforcing bar, and a background.

  At the construction site, the number of reinforcing bars, diameter, and pitch are measured by photographing the reinforcing bars of a reinforced concrete structure and processing the captured images with a computer (image processing). At that time, the rebar to be imaged (target rebar) usually overlaps with another rebar located behind and various backgrounds, so the image of the target rebar becomes unclear and the diameter measured as a result of image processing The accuracy such as length becomes worse. Therefore, in order to improve the accuracy of the measurement result by image processing, it is necessary to install a white board behind the object so that only the target reinforcing bar can be clearly photographed. Patent Document 1 discloses a bar arrangement information acquisition apparatus and method, and paragraph 0033 describes the installation of a white board.

JP 2010-122008

However, the bar arrangement information acquisition method of Patent Document 1 has the following problems.
(1) A white board cannot be inserted into a rebar with a small space unless its shape is devised. That is, a white board cannot be installed well behind the target reinforcing bar.
(2) It is necessary to affix a marker to the target reinforcing bar separately.
(3) It is difficult to measure both the main reinforcement and the reinforcement reinforcement with one white board.
(4) The shadow of the target rebar appears on the white board due to solar radiation or external illumination, and the rebar diameter may be measured incorrectly.

Therefore, in order to reduce the influence of the shadow at the time of shooting (to make the shadow whitish and to make the same group as the background bar by binarization), a black part was extracted as a reinforcing bar by the binarization method. However, the following problems occurred.
(1) When shooting outdoors, rebars and other shadows can be placed on the background bar due to the effects of solar radiation, and binarization cannot be performed accurately (specifically, rebars are extracted thicker than they actually are). (See FIGS. 22 (a) and (b)).
(2) When flash photography was performed, the reinforcing bars shined white and binarization could not be performed with accuracy (specifically, the reinforcing bars were extracted to be thinner than the actual ones) (FIGS. 22A and 22C). )reference).
(3) When the binarization cannot be performed with high accuracy, the accuracy may be improved by changing the binarization parameter. However, it is difficult to change the parameter unless it is a skilled user.
As described above, the method of extracting reinforcing bars by binarization using only the luminance threshold has a high false detection rate.

  The present invention has been made in view of the above problems, and a main object thereof is to easily and accurately acquire reinforcing bar arrangement information.

  In order to solve the above-described problem, the present invention provides a reinforcing bar region extracting device that extracts a reinforcing bar region from an image obtained by photographing a reinforcing bar, and in the image, pixels lined up perpendicular to the axial direction of the reinforcing bar. A first means for extracting the edge that is a boundary line of the region by connecting the places where the luminance is discontinuous between adjacent pixels, and based on the luminance distribution of each pixel in the image, A second means for calculating a standard luminance value, which is a standard value of luminance values in a reinforcing bar region, and a range of luminance in each region sandwiched between the extracted edges includes the calculated standard luminance value, And a third means for specifying the region as a region of a reinforcing bar.

  According to this configuration, by detecting a discontinuous portion of luminance in an image including a reinforcing bar, a reinforcing bar shadow, and a background, an edge that is a boundary line of the region is extracted, and a portion sandwiched between two edges is extracted. , Reinforcing bars or areas other than reinforcing bars. Since the region of the reinforcing bar is specified from the region using the standard luminance value based on the luminance distribution, erroneous detection of the reinforcing bar can be suppressed without binarizing the image. Since the image is not binarized, fine adjustment such as correction of the binarization parameter becomes unnecessary. According to the above, it is possible to easily and accurately extract a reinforcing bar region and acquire the reinforcing bar arrangement information.

Further, in the reinforcing bar region extracting device according to the present invention, the image includes a plurality of patterns provided at predetermined intervals in a direction perpendicular to the axial direction of the reinforcing bar, and the first means includes the first means, It is good also as scanning the brightness | luminance of a pixel in the direction based on the said pattern.
According to this configuration, when scanning the luminance of the pixels arranged perpendicular to the axial direction of the reinforcing bar in the image including the reinforcing bar, the direction to be scanned is determined by using the pattern included in the image. Can be easily identified.

In the reinforcing bar region extracting apparatus of the present invention, the first means performs a differential operation on the luminance of the adjacent pixel, and detects a pixel in which the luminance is discontinuous based on the calculated value. It is good as well.
According to this configuration, it is possible to accurately detect pixels whose luminance is discontinuous, and thus it is possible to accurately extract edges.

Further, in the reinforcing bar region extracting apparatus of the present invention, the first means sets the axial direction of the reinforcing bar as the vertical direction, and the vertical width of each of the extracted edges is equal to the width of the background bar, and An edge having an aspect ratio larger than a predetermined value may be set as a processing target of the third means.
According to this configuration, by leaving an edge having a vertical width equal to the width of the background bar and having a vertical length as an effective edge, it is possible to accurately grasp the region sandwiched between the edges. .

In the reinforcing bar region extracting apparatus of the present invention, the second means creates a frequency distribution of each luminance from the luminance distribution of each pixel in the image, and sets the minimum value among the luminance values at which the frequency is maximized. The standard luminance value in the reinforcing bar region may be used.
According to this configuration, an image including a reinforcing bar can be divided into three regions, that is, a reinforcing bar, a shadow of the reinforcing bar, and a background of the plate material, by edge, but the luminance of the reinforcing bar is the smallest among the three regions. Therefore, by setting the minimum value among the luminance values at which the frequency is maximized as the standard luminance value in the reinforcing bar region, it is possible to easily specify the reinforcing bar region from the region sandwiched between the edges.

Further, in the reinforcing bar region extracting apparatus of the present invention, when the specified reinforcing bar region is adjacent, the third means may integrate the adjacent regions and specify a single reinforcing bar region.
According to this configuration, when a region adjacent to one reinforcing bar region with an edge as a boundary line is not a reinforcing bar shadow or a background of a plate material but a separate reinforcing bar region (two reinforcing bar portions having different luminances). Are taken continuously). In such a case, it is very useful when calculating the diameter and pitch of the subsequent reinforcing bars by specifying the area of the integrated reinforcing bars rather than as separate reinforcing bar areas.

  The present invention also includes a reinforcing bar region extraction method and a reinforcing bar region extraction program. In addition, the problems disclosed by the present application and the solutions thereof will be clarified by the description of the mode for carrying out the invention and the drawings. The present invention is a further improvement of the invention of Japanese Patent Application No. 2010-211111 by the present applicant.

  According to the present invention, reinforcing bar arrangement information can be obtained easily and accurately.

It is a figure which shows the structure of the bar arrangement information acquisition system. 2 is a diagram showing a hardware configuration of a mobile terminal 4. FIG. 2 is a diagram illustrating a hardware configuration of a management server 5. FIG. It is a figure which shows the structure of the data memorize | stored in the bar arrangement information acquisition system 1, (a) shows the structure of the data memorize | stored in the memory | storage part 45 of the portable terminal 4, (b) is the memory | storage part of the management server 5 The structure of data stored in 55 is shown. It is a figure which defines the state of a deformed reinforcing bar, (a) shows the state of rib position 0 °, (b) shows the state of rib position 90 °, (c) shows the state of rib position 60 °. It is a figure which shows the structural example of the reinforcement standard information 452 of the portable terminal. It is a flowchart which shows the imaging | photography method of a rebar image. 3 is a flowchart illustrating a first embodiment of image processing by the portable terminal 4; It is a flowchart which shows the process of the reinforcement part extraction by the portable terminal. It is a flowchart which shows the reinforcement diameter estimation process by the portable terminal. It is a figure regarding the background bar B, (a) shows the shape of the background bar B, (b) shows the example of a marker. It is a figure which shows the example of an actual deformed reinforcing bar. It is a figure for demonstrating the acquisition processing of the main bar arrangement information, and shows the image which imaged the main bar and the background bar B. It is a figure for demonstrating the acquisition process of the main bar arrangement information, (a) shows the image which rotated the image of FIG. 13 only by -phi, (b) shows the figure for demonstrating a measurement principle. (A) shows an example of an image including a reinforcing bar and an extracted edge, (b) shows a case where a rectangle circumscribing the edge is vertically long, and (c) shows a case where a rectangle circumscribing the edge is not vertically long. . (A) shows the luminance distribution in the horizontal direction and the representative luminance values of the three regions at one pixel position in the axial direction (height direction) of the reinforcing bar, and (b) shows the extraction result of the reinforcing bar. It is a figure which shows the histogram of the brightness | luminance in an image, (a) shows the case where there is a shadow area | region, (b) shows the case where there is no shadow area | region. 6 is a flowchart showing a second embodiment of image processing by the portable terminal 4; It is a figure for demonstrating the acquisition processing of the reinforcement arrangement | positioning information of a main reinforcement and a reinforcement, (a) shows the image of the state which inserted the background bar b between the main reinforcement and the reinforcement, and (b) is ( An image obtained by binarizing only the portion of the background bar B in the image a) is shown. It is a figure for demonstrating the acquisition processing of the reinforcement arrangement | positioning information of a main reinforcement and a reinforcement reinforcement, and shows the image which rotated the recognized main reinforcement and the background bar B by -phi clockwise among the images of Fig.19 (a). . It is a figure for demonstrating the acquisition processing of the reinforcement arrangement | positioning information of a main reinforcement and a reinforcement, Comprising: The image which rotated the recognized reinforcement and the background bar | burr B clockwise by -phi among the images of Fig.19 (a). Show. (A) shows an image of an actual reinforcing bar and shadow, (b) shows a pattern in which the shadow is regarded as a reinforcing bar, and (c) shows a pattern in which the reinforcing bar is considered to be thin.

  Hereinafter, the best mode for carrying out the present invention will be described with reference to the drawings. In the bar arrangement information acquisition system according to the embodiment of the present invention, at a construction site, a background bar with markers (patterns) at both ends is arranged behind the deformed reinforcing bars, and the markers at both ends and a space between them are used using a digital camera. Take a picture of the deformed reinforcing bar, use the mobile terminal (rebar arrangement information acquisition device) to obtain the reinforcing bar information such as the number, diameter length, and pitch (interval) of the deformed reinforcing bar from the photographed image, and type of each reinforcing bar from the distribution of diameter length (Standard, nominal diameter and nominal diameter).

  In particular, it is assumed that the flash is not used when photographing reinforcing bars, and the edge that is the boundary line of the region is detected based on the brightness change value in the captured image including the reinforcing bar, and the luminance in each region sandwiched between the edges is detected. And a representative luminance value (standard luminance value) that is a representative value (standard value) of luminance values in the reinforcing bar region calculated from the luminance distribution of the image, the reinforcing bar region is specified. In addition, by making the background bar into a shape that can be inserted between the reinforcing bars, and inserting the background bar diagonally, the diameters of the main bars and shear reinforcing bars (hereinafter simply referred to as “reinforcing bars”) are measured simultaneously. can do. In the present embodiment, the main bars are arranged in the vertical direction, and the reinforcing bars are arranged in the horizontal direction.

  According to this, the false detection rate of the reinforcing bar portion can be greatly reduced. Further, since arrangement information can be obtained easily and accurately at the site, the correctness of the completed shape can be judged on the spot by comparing and collating with design drawing information.

≪System configuration and overview≫
FIG. 1 is a diagram illustrating a configuration of a bar arrangement information acquisition system 1. The bar arrangement information acquisition system 1 includes a digital camera 3 and a portable terminal 4 at a construction site, and a management server 5 at an office. Data can be transmitted / received between the digital camera 3 and the portable terminal 4 through a USB (Universal Serial Bus) cable or the like. Data can be transmitted and received between the portable terminal 4 and the management server 5 by wireless communication or the like.

  The digital camera 3 is used to take an image of the imaging target part 2 such as a pillar, a beam, a floor, or a wall including a reinforcing bar, and has a pixel count of, for example, 4 million pixels or more and can turn off the autofocus function. It is done. The portable terminal 4 is a portable information processing device, takes in a digital image taken from the digital camera 3 to generate reinforcement information, receives design drawing information from the management server 5, and arranges information and design drawing information. Are compared and checked to determine whether the finished shape is valid. The mobile terminal 4 may be replaced by a PC (Personal Computer) or a server. The management server 5 includes a storage unit 55 that stores design drawing information and construction photo information, and transmits and receives such information to and from the mobile terminal 4.

<< Device configuration >>
FIG. 2 is a diagram illustrating a hardware configuration of the mobile terminal 4. The portable terminal 4 includes a communication unit 41, a display unit 42, an input unit 43, a processing unit 44, and a storage unit 45. The communication unit 41 is a part that performs data communication with the digital camera 3 and the management server 5, and is realized by, for example, a USB port or a NIC (Network Interface Card). The display unit 42 is a part that displays data according to an instruction from the processing unit 44, and is realized by, for example, a liquid crystal display (LCD). The input unit 43 is a part where an operator inputs data (for example, data such as reinforcing bar standard information), and is realized by, for example, a keyboard or a mouse. The processing unit 44 delivers data between each unit and controls the entire mobile terminal 4 and is realized by a CPU (Central Processing Unit) executing a program stored in a predetermined memory. The The storage unit 45 stores data from the processing unit 44 and reads the stored data, and is realized by, for example, a nonvolatile storage device such as a flash memory or a hard disk device.

  FIG. 3 is a diagram illustrating a hardware configuration of the management server 5. The management server 5 includes a communication unit 51, a display unit 52, an input unit 53, a processing unit 54, and a storage unit 55. The communication unit 51 is a part that performs data communication with the portable terminal 4 via a wireless network, and is realized by, for example, a NIC or the like. The display unit 52 is a part that displays data in accordance with an instruction from the processing unit 54, and is realized by, for example, a liquid crystal display. The input unit 53 is a part where an operator inputs data (for example, data such as design drawing information), and is realized by, for example, a keyboard or a mouse. The processing unit 54 exchanges data between the units and controls the entire management server 5 and is realized by the CPU executing a program stored in a predetermined memory. The storage unit 55 stores data from the processing unit 54 and reads the stored data, and is realized by, for example, a nonvolatile storage device such as a flash memory or a hard disk device.

<< Data structure >>
FIG. 4 is a diagram illustrating a configuration of data stored in the bar arrangement information acquisition system 1. FIG. 4A shows a configuration of data stored in the storage unit 45 of the mobile terminal 4. The storage unit 45 stores an image processing program 451 and reinforcing bar standard information 452. The image processing program 451 is a program that acquires reinforcement information from image data captured by the digital camera 3 and performs a process of determining compatibility with the design drawing information. The processing unit 44 according to the necessity of the process. Is read from the storage unit 45 in response to the instruction. Reinforcing bar standard information (reinforcing bar type information) 452 is table information used for obtaining a reinforcing bar standard (type) from the distribution of diameters. Details thereof will be described separately.

  FIG. 4B shows a configuration of data stored in the storage unit 55 of the management server 5. The storage unit 55 stores design drawing information 551 and construction photograph information 552 in advance. The design drawing information 551 is drawing information related to the design of a building such as a reinforcing bar (including the diameter of the deformed reinforcing bar), registered in the storage unit 55 by the administrator, and from the management server 5 to the portable terminal 4 as necessary. Sent to. The construction photo information 552 is photo information of a building at an actual construction site. Photo information taken by the digital camera 3 is transmitted to the management server 5 via the mobile terminal 4 and stored in the storage unit 55.

  FIG. 5 is a diagram for defining the state of the deformed reinforcing bar. The deformed reinforcing bar is one of building structural materials, and is a bar-shaped steel material in which an iron bar is rolled to provide unevenness on the surface. As the unevenness, as shown in FIG. 5, nodes (ribs) and ribs are provided. When viewed (taken) from a direction perpendicular to the axis of the reinforcing bar, the diameter of the deformed reinforcing bar varies depending on the position (angle) of the rib. Hereinafter, the three states will be described with the rib facing the front as a rib position of 0 °.

FIG. 5A shows a state where the rib position is 0 °. Since the rib in this state is located on the front surface, the radial length is not affected. The right and left nodes as viewed in the drawing are alternately provided along the axial direction. Accordingly, as the diameter length of the deformed reinforcing bar, a diameter length d0 not including a node and a diameter length d1 including one node are extracted.
FIG. 5B shows a state where the rib position is 90 °. Since the ribs in this state are located at both ends and the nodes are included in the ribs, the diameter length d2 including the ribs at both ends is extracted as the diameter length of the deformed reinforcing bar.
FIG. 5C shows a state where the rib position is 60 °. The rib in this state affects the diameter length depending on the height of the protrusion. You can see the right-hand section but not the left-hand section when looking at the drawing. Therefore, as the diameter length of the deformed reinforcing bar, a diameter length d3 not including a node and a diameter length d4 including one node are extracted. In addition, FIG. 12 is a figure which shows the example of an actual deformed reinforcing bar.

  FIG. 6 is a diagram illustrating a configuration example of the reinforcing bar standard information 452 of the mobile terminal 4. The reinforcing bar standard information 452 is table information for estimating the rib position from the distribution of the diameter length in the photographed reinforcing bar image, and further specifying the standard (type) of the corresponding reinforcing bar. The nominal diameter 4521, the nominal diameter 4522, and the rib It is composed of records including a position 4523. A nominal diameter 4521 indicates the nominal diameter of the deformed reinforcing bar. The nominal diameter 4522 indicates a diameter (diameter length) generally referred to for a deformed reinforcing bar having a nominal diameter 4521. The rib position 4523 indicates the position (angle) of the rib when the front face of the rib is 0 °, and the rib position 4523 is 0 to 60 °, 60 to 75 °, and 75 to 90 °. The lower limit value and the upper limit value of the diameter length are shown for each of the two cases. Since the appearance of the diameter varies depending on the shape and size of the nodes and ribs, the range of the rib position 4523 may be divided into two, or may be divided into four or more.

  In the estimation of the diameter of the reinforcing bar, first, one of the three rib positions 4523 is estimated from the distribution of the lengths, and the median of the lengths is compared with the range of the lower limit value and the upper limit value. Reinforcing bar standards with diameter 4521 and nominal diameter 4522 are specified. Details will be described later.

≪System processing≫
FIG. 7 is a flowchart illustrating a method for capturing a reinforcing bar image. In the construction site, a photographer photographs a reinforcing bar using the background bar and the digital camera 3, transfers the photographed image to the portable terminal 4, and acquires reinforcement information from the photographed image using the portable terminal 4. The procedure is shown.

  First, the photographer determines the digital camera 3 (S701). At that time, it is confirmed whether or not the setting of 4 million pixels or more is possible and the autofocus function is cut off (S702). If the condition is not met (NO in S702), the digital camera 3 is selected again. Correct (S701). If the condition is met (YES in S702), the determined autofocus function of the digital camera 3 is turned off so that the focus is achieved at an appropriate distance (for example, 2 m) between the imaging target region 2 and the digital camera 3. (S703). Thereafter, the focal length is constant. Then, the calibration board is photographed and camera parameters are acquired (S704). This is called camera calibration, and detects distortion or the like of the digital camera 3 by photographing a board which is paper on which a lattice pattern or equally spaced dots are printed.

  Next, the photographer determines an imaging target region 2 including a reinforcing bar (S705), and inserts a background bar behind the reinforcing bar (S706). In order to fix the background bar, it is conceivable to use a magnet or the like when only the main bar is present. On the other hand, when both the main bar and the reinforcing bar are present, it may be leaned between the intricate reinforcing bars.

  FIG. 11A shows the shape of the background bar B. FIG. The background bar B includes a plate material BA and markers MK1 and MK2. The plate material BA is a strip-shaped plate material that is used as a background when photographing the reinforcing bar, and a color (for example, white) different from that of the reinforcing bar is attached to at least one surface.

Width _{T 1} of the plate BA, the pitch in the section deformed bars (e.g., 20 mm) from the large and spacing of reinforcement (e.g., 100 mm) smaller than, for example, is formed on 50 mm. On the other hand, the marker MK1 and MK2 is subjected to both ends of the plate BA, the width _{T 2} are, spacing reinforcement (e.g., 100 mm) smaller than, for example, is formed on 70mm or less.

According to this, the width T ₁ of the plate BA is larger than the pitch of the knots, regardless of the position of the background bars B, can be photographed in a state in which any of the knots is always present in front of the background bars B. Therefore, the maximum diameter length of the reinforcing bar can be measured while grasping the distribution of fushi. Then, since the size of the width T ₁ is limited, thereby shortening the calculation time. Further, since the width T ₂ of the width T ₁ and the marker MK1, MK2 plate member BA is smaller than the pitch of the reinforcement, without interfering with the reinforcement, readily background bar B can be inserted into the reinforcing bars.

  A marker is a characteristic shape that does not exist in nature, and its size (dimension) is known in advance, and it is per pixel from the size of the marker and the number of pixels in the marker in the captured image. , While recognizing the reinforcing bar between them by scanning between the two markers MK1 and MK2, and, further, between the imaging target region 2 and the digital camera 3 Used to estimate the distance between.

  FIG. 11B is a diagram illustrating an example of a marker. Cross markers and circular markers are shown. Since markers MK1 and MK2 are assigned to the background bar B and the background bar B is placed behind the reinforcing bar, the marker and the reinforcing bar at the same distance from the digital camera 3 can be photographed. Since the length per pixel is equal between the steel bar and the reinforcing bar, the diameter length and pitch can be obtained with high accuracy.

  Then, the object is photographed by the digital camera 3 from a position separated by an appropriate distance (for example, 2 m) (S707), the photographed image is transferred to the portable terminal 4, and image processing is executed (S708). The vertical angle in the shooting direction of the digital camera 3 is about 0 °, and the horizontal angle is arbitrary. Details of the image processing by the portable terminal 4 will be described later. After the image processing, the calculation result is displayed on the display unit 42 of the portable terminal 4 (S709). Then, if the calculation result is the reinforcing bar diameter and pitch of the design drawing information (YES in S710), it is determined that the bar arrangement state is normal, and the photographing operation is terminated. On the other hand, if the calculation result is not the reinforcing bar diameter and pitch of the design drawing information (NO in S710), the photographer instructs the worker on the site to correct the bar arrangement (S711), and after the correction is performed, the photograph is taken. The target region 2 is determined again (S705), and imaging work is performed.

  FIG. 8 is a flowchart showing a first embodiment of image processing by the portable terminal 4. This process is performed when the portable terminal 4 acquires image data from the digital camera 3 and stores it in a built-in image memory. In this embodiment, processing when the imaging target region 2 is only the main muscle will be described.

  First, the portable terminal 4 calls an image processing program (S801). Specifically, the processing unit 54 reads the image processing program 451 from the storage unit 55, loads it into the main storage device (main memory), and positions the program counter (control pointer) at the start address of the image processing program 451. As a result, the processing unit 44 of the portable terminal 4 starts processing according to the image processing program 451. The processing flow is shown in S802 to S814.

  First, the portable terminal 4 (processing unit 44) acquires a digital image, a camera focal length FL [pixel], a reference length [mm] within the markers MK1 and MK2, and an inter-marker distance M [mm] (S802). A digital image is acquired from the digital camera 3 in the image memory via a USB cable. The camera focal length FL, the in-marker reference length, and the inter-marker distance M are acquired through the input unit 53 by a photographer's operation. The in-marker reference length is a reference length for the markers MK1 and MK2. For example, in the case of a circular marker, the length of the diameter of the circle is applied. Next, image correction and binarization are performed (S803). More specifically, the distortion or the like of the digital image is corrected using the camera parameters acquired in S704, and the corrected digital image is converted into a monochrome image with a pixel value = 0 (black) or 1 (white).

  Subsequently, the mobile terminal 4 measures, from the binarized image data, the angle formed by the main line S with the vertical line and the angle formed by the background bar B with the horizontal line (S804). Referring to FIG. 13, a vertical line means a straight line parallel to the vertical direction of the pixel arrangement in the image memory IM, and a horizontal line means a straight line parallel to the horizontal direction of the pixel arrangement in the image memory IM. To do. The angle is positive in the clockwise direction. A rectangular region having a pixel value of 0 (black) extending in the vertical direction in the image memory IM is defined as a main line S, and a straight line parallel to the long side of the rectangle and passing through the center of the rectangle is a center line of the main line S. And the angle between the center line and the vertical line is θ. Further, a straight line passing through the centers of the markers MK1 and MK2 is extracted as the center line of the background bar B, and the angle between the center line and the horizontal line is φ.

  Subsequently, the mobile terminal 4 rotates the original image clockwise by −φ, and makes the background bar B parallel to the horizontal line (S805). FIG. 14A shows the image of FIG. 13 rotated by −φ, and the center line of the background bar B is parallel to the horizontal line.

Next, an image of the reinforcing bar portion is extracted (S806). In the upper part of FIG. 14B, an image obtained by extracting the marker and the reinforcing bar portion of the background bar B is shown. Then, the maximum number of pixels along the background bar B of the markers MK1 and MK2 is counted (S807). Specifically, the maximum diameter among the circular diameters that are markers is specified, and the number of pixels included in the diameter is counted. Then, the length per pixel is obtained by dividing the reference length in the marker by the number of pixels. According to this, by using a circular marker, the maximum diameter is constant no matter which direction the marker is viewed. Therefore, even if the orientation of the marker changes, the length per pixel can be accurately specified. . Further, distances L _m1 , L _m2 and L _H from the camera are obtained. (S808). As shown in FIG. 14B, the distance L _m1 is a distance between the camera and the marker MK1. The distance L _m2 is a distance between the camera and the marker MK2. The distance L _H is a camera, the shortest distance between the background bar B, used in calculating the camera, the distance between each reinforcing bar. Each distance is calculated | required by the following formula | equation 1,2,3.
L _m1 = FL × (reference length of marker MK1 / number of pixels of marker MK1) Equation 1
L _m2 = FL × (reference length of marker MK2 / number of pixels of marker MK2) Equation 2
L _H = √ (L _m1 ² − (L _m1 ² −L _m2 ² + M ² ) ² / 4M ² ) Equation 3

  Furthermore, the portable terminal 4 estimates the diameter of each main muscle (S809). By estimating the diameter of the main bar, the type of the reinforcing bar is specified. Details of this processing will be separately described as subroutine processing.

  Next, the portable terminal 4 acquires the coordinates (position), posture, and design drawing information of the digital camera 3 (S810). The coordinates of the digital camera 3 are acquired as position information given to the captured image by connecting a GPS (Global Positioning System) device, for example. The attitude of the digital camera 3 is acquired by a function (function of the digital camera 3 or connected device) that detects the camera attitude at the time of shooting. The design drawing information is acquired when the mobile terminal 4 receives the design drawing information 551 stored in the storage unit 55 of the management server 5. Then, the target part (construction part) of the photographed image is specified, and the compatibility with the corresponding design drawing information is determined (S811). For example, the compatibility between the estimated deformed reinforcing bar diameter and the deformed reinforcing bar diameter included in the design drawing information is determined.

  If the drawing is completed (YES in S812), the portable terminal 4 outputs the design drawing information and the recognition information to the display unit 42 (S813). The design drawing information is a target portion, coordinates, number, pitch, diameter length, etc. of the reinforcing bar on the drawing. The recognition information is the actual number of reinforcing bars, pitch, and diameter. If the drawing is not made (NO in S812), the arrangement abnormality abnormality information indicating the abnormality content, the design drawing information, and the recognition information are output to the display unit 42 (S814). Note that the compatibility determination result can be transmitted to another device through the communication unit 41 instead of being output to the display unit 42.

  FIG. 9 is a flowchart illustrating a reinforcing bar portion extraction process performed by the mobile terminal 4. This is the processing of the reinforcing bar portion extraction subroutine in the image processing program, where edges (region boundaries) are extracted from the image of the background bar portion, and each portion sandwiched between two edges has three regions. By discriminating one of the above, a reinforcing bar part is specified as one of those areas.

  First, the mobile terminal 4 binarizes the entire image corresponding to the portion of the background bar B, and detects the markers MK1 and MK2 (S901). Next, an image between the markers MK1 and MK2 is cut out (S902). At this time, it is assumed that the clipped image is not binarized. The upper figure of Fig.15 (a) shows the example of the image cut out.

  Then, the mobile terminal 4 extracts an edge from the cut image (S903). An edge is a boundary line between three regions (rebar, shadow, background) in an image. For example, when the luminance of pixels arranged in the image perpendicular to the axis direction of the reinforcing bar is scanned, a portion where the luminance becomes discontinuous (changes rapidly) becomes an edge. Therefore, for example, the luminance of pixels adjacent in the scan direction is subjected to Laplacian (second order differential) filter processing, and the absolute value of the calculated value is a predetermined value or more (specifically, for example, the middle of two adjacent pixels) ), And a line connecting the detected portions is extracted as an edge. For example, when the markers MK1 and MK2 are detected in the image of the background bar B shown in FIG. 11A, the direction in which the luminance of the pixel is scanned is the direction from the markers MK1 to MK2 (the longitudinal direction of the background bar B). ). The lower part of FIG. 15A shows an example of the extracted edge. It should be noted that a primary differential filter process can also be used when detecting a location where the luminance is discontinuous.

  Subsequently, the mobile terminal 4 removes unnecessary edges from the extracted edges (S904). For example, assuming that the axial direction of the reinforcing bar is the vertical direction, a rectangle circumscribing each edge is drawn, and the vertical width and aspect ratio of the rectangle are calculated. Then, the drawn rectangle has a vertical width equal to the width of the background bar B, and as shown in FIG. 15B, an edge is left where the rectangle is vertically long (aspect ratio> α). On the other hand, when the vertical width of the rectangle is smaller than the width of the background bar B, or as shown in FIG. 15C, the rectangle is not vertically long (aspect ratio ≦ α), that is, square or horizontally long. The edge is excluded. The aspect ratio threshold value α for determining whether or not the image is vertically long is a predetermined value of 1 or more.

  Next, the mobile terminal 4 divides the portion between each edge into three regions (rebar, shadow, background) (S905). For example, assume that a luminance distribution in a direction perpendicular to the axial direction of the reinforcing bar, indicated by A in FIG. This luminance distribution is obtained by scanning the luminance of pixels from the left side to the right side of the background. A histogram (frequency distribution) is created for the luminance distribution of the pixels included in the entire image, and the maximum value (the luminance value at which the frequency is maximized) is labeled. Then, the representative luminance value and the intermediate luminance value of each region are calculated from the labeled maximum values.

  Hereinafter, a procedure for calculating the representative luminance values of the three areas based on the luminance distribution of the image and identifying which area is the portion sandwiched between the edges based on the representative luminance values will be described.

  FIG. 17 is a diagram showing a histogram of image luminance, in which the horizontal axis indicates luminance and the vertical axis indicates frequency. FIG. 17A shows a case where there is a shadow area. There are six local maximum values, and they are labeled as min1, min2,... In order from the lowest luminance, and are labeled as max1, max2,. The maximum value max1 is the maximum value with the maximum luminance, and becomes the representative luminance value B of the background. The local maximum value min1 is the local maximum value with the minimum luminance, and becomes the representative luminance value C of the reinforcing bar. Further, an intermediate luminance value between the maximum value max1 and the maximum value max2 is set as a maximum value mid_max. Next, an intermediate luminance value between the maximum value min1 and the maximum value min2 is set as a maximum value mid_min. At this time, the intermediate luminance value between the maximum value mid_max and the maximum value mid_min becomes the representative luminance value D of the shadow.

  FIG. 17B shows a case where there is no shadow area. If there is no shadow area, the background immediately follows the reinforcing bar. There are two maximum values, and the one with the smaller luminance is labeled as the maximum value min1, while the one with the larger luminance is labeled as the maximum value max1. The maximum value max1 is the background representative luminance value B. The maximum value min1 becomes the representative luminance value C of the reinforcing bar. When there are only two local maxima, it is considered that there is no shadow area.

  As shown in FIG. 16A, the representative luminance value B of the background, the representative luminance value C of the reinforcing bar, and the representative luminance value D of the shadow are specified from the histogram of FIG. Therefore, the region between the edges is specified as one of reinforcing bar, shadow, and background based on the representative luminance value. More specifically, since the luminance range in each region sandwiched between the edges is determined, depending on whether the representative luminance value B, C, or D is included in the range, the background, the reinforcing bar, or the shadow is the region. As specified. For example, since the luminance range of the region E includes the representative luminance value C, it can be seen that the region is a reinforcing bar portion.

  Furthermore, when the adjacent areas are the same among the specified areas, the mobile terminal 4 integrates the areas (S906). For example, if there are three edges, two regions are formed between the edges, but if the two regions are both reinforcing bars, the central edge is eliminated and the region is sandwiched between the edges at both ends. Is one reinforcing bar region.

  As a result of the above processing, reinforcing bars are extracted as shown in FIG. Although not shown in FIG. 16B, a reinforcing bar may come after the reinforcing bar and shadow instead of the background (that is, a reinforcing bar, a shadow, and a reinforcing bar).

  FIG. 10 is a flowchart showing a rebar diameter estimation process by the mobile terminal 4. This is a processing of a rebar diameter estimation subroutine including main and reinforcing bars in the image processing program. The continuous diameter of each reinforcing bar is extracted from the binarized image data, and the diameter length data is extracted. The number of data and the mode value are obtained by shaping, and the standard of the reinforcing bar is specified according to the value obtained by dividing the mode value by the number (index value of variation in diameter length data).

  First, the portable terminal 4 extracts a continuous diameter [mm] corresponding to the position of 1 pixel in the vertical direction with respect to the background bar B (S1001). 14B, first, a pixel having a pixel value of 0 (black) is searched from a marker MK1 to MK2 at a predetermined position in a direction perpendicular to the background bar B, and along the background bar B having a reinforcing bar diameter. Obtain the number of pixels Wp. At that time, the number P of pixels along the background bar B between the reinforcing bars and the markers is also obtained. Here, the number of pixels between reinforcing bars, that is, the bar arrangement pitch, indicates the interval between the central axes of adjacent reinforcing bars, and is calculated as an average value of the number of pixels between the left edges and the number of pixels between the right edges, for example. The

Subsequently, the distance m along the background bar B between the reinforcing bar and the marker is obtained, the distance Lt from the camera to each reinforcing bar is obtained, and the reinforcing bar diameter Wm and the reinforcing bar pitch m ′ are obtained. For example, the reinforcing bar 2 in FIG. 14B can be obtained by the following equations 4, 5, 6, and 7. In addition, Formula 6 and Formula 7 are the calculation formulas regarding the main muscle. The reinforcement bars will be described later.
m ₂ = M × P ₂ / ΣPn (ΣPn: n = 1 to 6) Expression 4
L _t2 = √ (L _H ² + m ₂ ² ) Equation 5
W _m2 = W _P2 × L _t2 / FL × cos (φ + θ) Equation 6
m ₂ ′ = m ₂ × cos (φ + θ) Equation 7
When _obtaining L _t1 , P ₂ in Equation 4 is replaced with P ₁ + P ₂ . When _calculating L _t3 , P ₂ in Equation 4 is replaced with P ₃ . When _calculating L _t4 , P ₂ in Equation ₄ is replaced with P ₃ + P ₄ .

  Next, as the data shaping, the mobile terminal 4 extracts a diameter corresponding to, for example, a position of ± 500 pixels from the center line of the background bar B in a direction perpendicular to the background bar B (S1002). Then, the median value M of 1000 extracted diameters is acquired (S1003). In this case, the median value M is an average value of the 500th diameter length and the 501st diameter length. Further, as data shaping, data corresponding to 0.8M to 1.2M is extracted from 1000 diameter length data (S1004). Then, the number N of the extracted diameter length data is acquired (S1005). Further, as data shaping, data less than N × 0.01 is deleted (S1006). According to this, since the number of data less than 1% of the number N of diameter data is excluded, extremely large or extremely small diameter data caused by a measurement error or the like can be excluded. . Thereafter, the number of data Ni and the mode value Ymax after shaping are acquired (S1007).

  Therefore, the value of Ymax / Ni is obtained, and the state of the reinforcing bar (rib position) is estimated according to the value. This is because deformed reinforcing bars have nodes and ribs, so the appearance of the nodes and ribs changes depending on the direction of the reinforcing bars, and the distribution of the extracted diameter data also changes. The direction of the reinforcing bar is estimated from the index value.

  First, when the value of Ymax / Ni is less than 0.35 (YES in S1008), the reinforcing bar state (rib position) is 0 ° to 60 ° (S1009), and the median Mi after data shaping is acquired ( S1010), the field of 0-60 is referred among the rib positions 4523 in the reinforcing bar standard information 452 (see FIG. 6) of the storage unit 45 (S1011). When the value of Ymax / Ni is 0.35 or more and less than 0.45 (YES in S1012), the reinforcing bar state (rib position) is 60 ° to 75 ° (S1013), and the median after data shaping Mi is acquired (S1014), and the fields 60 to 75 in the rib position 4523 in the reinforcing bar standard information 452 of the storage unit 45 are referred to (S1015). When the value of Ymax / Ni is 0.45 or more (NO in S1012), the reinforcing bar state (rib position) is 75 ° to 90 ° (S1016), and the median Mi after data shaping is acquired (S1017). In the rib position 4523 in the reinforcing bar standard information 452 of the storage unit 45, the column 75 to 90 is referred to (S1018).

  As a result of referring to the reinforcing bar standard information 452, it is determined whether there is a reinforcing bar standard (nominal diameter 4521 and nominal diameter 4522) corresponding to the median value Mi (S1019). Specifically, it is determined whether or not a combination of a lower limit value and an upper limit value of the diameter length including the median value Mi is in each column. If it exists (YES in S1019), the corresponding reinforcing bar standard is acquired. If it does not exist (NO in S1019), the measurement has failed (S1021). According to this, since the reinforcing bar standard information 452 corresponding to the rib position (orientation) of the reinforcing bar is used, the reinforcing bar standard can be acquired with high accuracy regardless of the angle taken from any angle.

≪Processing when main bars and reinforcement bars coexist≫
FIG. 18 is a flowchart showing a second embodiment of the image processing by the portable terminal 4, and the processing of S1805 and S1811 to S1815 is added to the flowchart of FIG. In this embodiment, a description will be given of a process of discriminating between a main muscle and a reinforcing bar and acquiring respective bar arrangement information when the main bar and the reinforcing bar are mixed in the imaging target region 2.

  First, the portable terminal 4 calls an image processing program, acquires a digital image, a focal length, and the like, and binarizes the image, similarly to the processing of S801 to S804 in FIG. The angle θ is measured (S1801 to S1804). Subsequently, the pixel value binarized in S1803 is rotated clockwise by −θ so that the center line of the main muscle S is parallel to the vertical line, and only the portion of the background bar B is extracted. It is determined whether the area is 0 (black), which is the main reinforcing bar or the reinforcing bar (S1805). At this time, the number of reinforcing bars is known from the number of black areas.

  FIGS. 19-21 is a figure for demonstrating a process when a main reinforcement and a reinforcement reinforcement are mixed. FIG. 19A shows an image in which the background bar B is inserted diagonally between the main and reinforcing bars, the angle formed by the main and reinforcing bars is 90 °, and the center line of the main bar S is a vertical line. The angle between the center line of the background bar B and the horizontal line is φ. Next, FIG. 19B shows an image in which only the background bar B is extracted from the image of FIG. In the area of the image where the pixel value is 0 (black), the location where the angle between the center line and the horizontal line is 90 ° is recognized as the main muscle, and Wspi (i = 1 to 5) is the diameter of the main muscle. On the other hand, a portion where the angle between the center line and the horizontal line is 0 ° is recognized as a reinforcing bar, and Whpj (j = 1, 2) is the diameter of the reinforcing bar.

  Next, the mobile terminal 4 rotates the main bar and the background bar B by −φ so that the center line of the background bar B is parallel to the horizontal line (S1806). FIG. 20 shows an image obtained by rotating the recognized main streak and the background bar B by −φ clockwise among the images in FIG. Then, if an image of only the background bar B is extracted (S1807), the image of the marker and the main streak as shown in the upper part of FIG. 14B is obtained, which is the same as the processing of S807 to S809 in FIG. The diameter length and pitch of each main muscle are calculated (S1808 to S1810). Specifically, the calculation is performed according to Equations 1 to 7. In Equations 6 and 7, cos φ is used instead of cos (φ + θ).

  Subsequently, the mobile terminal 4 rotates the reinforcing bars and the background bar B by −φ so that the center line of the background bar B is parallel to the horizontal line (S1811). FIG. 21 shows an image obtained by rotating the recognized reinforcing bars and the background bar B by −φ clockwise among the images in FIG. Then, if an image of only the background bar B is extracted (S1812), it becomes an image of the marker and the reinforcing bar as shown in the upper part of FIG. 14B, so that it is the same as the processing of S807 to S809 of FIG. Then, the diameter length and pitch of each reinforcing bar are calculated (S1813 to S1815). Specifically, the calculation is performed according to Expressions 1 to 7. In Expressions 6 and 7, sin φ is used instead of cos (φ + θ).

  Furthermore, the portable terminal 4 determines whether or not the calculated bar arrangement information is in accordance with the design drawing information, and outputs information according to the result (S1816 to S1820), similarly to the processing of S810 to S814 in FIG. ).

  Although the embodiment of the present invention has been described above, in order to make each device of the bar arrangement information acquisition system 1 shown in FIG. 1 function, the program executed by the processing unit of each device is stored on a computer-readable recording medium. It is assumed that the bar arrangement information acquisition system 1 according to the embodiment of the present invention is realized by recording, causing the computer to read and execute the recorded program. Note that the program may be provided to the computer via a network such as the Internet, or a semiconductor chip or the like in which the program is written may be incorporated in the computer.

  According to the embodiment of the present invention described above, first, the background bar B in which the markers MK1 and MK2 and the white plate material BA are integrated may be inserted into the reinforcing bar. There is no need for a ruler with a scale, and there is less to set before shooting. Next, by using one background bar B, it is possible to easily obtain the arrangement information of both the main reinforcement and the reinforcement.

  Then, an edge is extracted by detecting a discontinuous portion of luminance in a photographed image without using a flash, and a reinforcing bar is extracted from a region sandwiched between the edges based on the representative luminance value calculated from the luminance distribution. Thus, the false detection rate of the reinforcing bar portion can be greatly reduced. Then, by extracting the shadow portion, it is possible to further eliminate false detection. Further, fine adjustment such as binarization parameter correction is not required, and the reinforcing bar portion can be extracted by a simple program.

<< Other embodiments >>
Although the best mode for carrying out the present invention has been described above, the above embodiment is intended to facilitate understanding of the present invention and is not intended to limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and equivalents thereof are also included in the present invention. For example, the following embodiments can be considered.

(1) In the above embodiment, the median value is acquired from the 1000 diameter length data in S1003 of FIG. 10, but other representative values (for example, the mode value) are acquired. Also good.

(2) The marker is not limited to a circle, but may be a square when viewed from the front. Further, the number of markers is not limited to two at both ends of the plate material BA, but may be three or more at predetermined intervals in the longitudinal direction of the plate material BA. At this time, it is conceivable to select a reinforcing bar for which bar arrangement information needs to be acquired, specify two markers at a position sandwiching the selected reinforcing bar from a plurality of markers, and use the two markers.

(3) The angle formed by the main line S with the vertical line and the angle formed by the background bar B with the horizontal line were measured, and these angles were used, but not limited to the vertical line and the horizontal line. If the angle with respect to the direction or the lateral direction is fixed, another reference line may be used, and not only the two reference lines for the main muscle S and the background bar B, but also one reference line may be used. .

(4) In the above embodiment, for example, as illustrated in FIG. 14B, the description has been made on the assumption that the camera and the background bar are facing each other (the optical axis of the camera and the background bar are orthogonal). In addition, when the camera and the reinforcing bar and the background bar are not facing each other, for example, when the background bar is inclined in the depth direction with respect to the camera, the diameter length of the reinforcing bar is determined by the same formulas 1 to 7 as in the above embodiment. It is possible to measure.

DESCRIPTION OF SYMBOLS 1 Reinforcement information acquisition system 2 Imaging object part 3 Digital camera 4 Portable terminal (reinforcement information acquisition apparatus, information processing apparatus)
44 processing unit 45 storage unit 451 rebar standard information (rebar type information)
5 Management server B Background bar (rebar shooting tool)
BA plate material FL Focal length Lm, L _H , Lt Distance from camera MK1, MK2 Marker (pattern)
Wp pixels

Claims (8)

A rebar area extraction device that extracts a rebar area from an image of a rebar,
In the image, the luminance of pixels arranged perpendicular to the axial direction of the reinforcing bar is scanned, and the locations where the luminance is discontinuous between adjacent pixels are connected to extract an edge that is a boundary line of the region. A first means;
A second means for calculating a standard luminance value that is a standard value of the luminance value in the region of the reinforcing bar based on the luminance distribution of each pixel in the image;
A third means for specifying the region as a reinforcing bar region when the luminance range in each region sandwiched between the extracted edges includes the calculated standard luminance value;
A reinforcing bar region extracting apparatus comprising: The reinforcing bar region extracting device according to claim 1,
The image includes a plurality of patterns provided at predetermined intervals in a direction perpendicular to the axial direction of the reinforcing bar,
The first means scans the luminance of the pixel in a direction based on the pattern. The reinforcing bar region extracting device according to claim 1,
The first means performs a differential operation on the luminance of the adjacent pixels, and detects pixels in which the luminance is discontinuous based on the calculated value. The reinforcing bar region extracting device according to claim 1,
The first means has an axial direction of the reinforcing bar as a vertical direction, and among the extracted edges, an edge whose vertical width is equal to a width of a background bar and whose aspect ratio is larger than a predetermined value, Reinforcing bar region extraction apparatus, which is a processing target of the third means. The reinforcing bar region extracting device according to claim 1,
The second means creates a frequency distribution of each luminance from the luminance distribution of each pixel in the image, and sets the minimum value among the luminance values at which the frequency is maximized as the standard luminance value in the reinforcing bar region. Reinforcing bar region extraction device. The reinforcing bar region extracting device according to claim 1,
When the specified reinforcing bar areas are adjacent to each other, the third means integrates the adjacent areas and specifies one reinforcing bar area. The reinforcing bar area extracting apparatus. A reinforcing bar region extraction method for extracting a reinforcing bar region from an image obtained by photographing a reinforcing bar by an information processing device,
The information processing apparatus includes:
In the image, the luminance of pixels arranged perpendicular to the axial direction of the reinforcing bar is scanned, and the locations where the luminance is discontinuous between adjacent pixels are connected to extract an edge that is a boundary line of the region. The first step;
A second step of calculating a standard luminance value which is a standard value of the luminance value in the region of the reinforcing bar based on the luminance distribution of each pixel in the image;
A third step of specifying the region as a reinforcing bar region when the luminance range in each region sandwiched between the extracted edges includes the calculated standard luminance value;
Reinforcement region extraction method characterized by performing A reinforcing bar region extraction program for causing an information processing device to execute a method of extracting a reinforcing bar region from an image of a reinforcing bar,
In the image, the luminance of pixels arranged perpendicular to the axial direction of the reinforcing bar is scanned, and the locations where the luminance is discontinuous between adjacent pixels are connected to extract an edge that is a boundary line of the region. The first step;
A second step of calculating a standard luminance value which is a standard value of the luminance value in the region of the reinforcing bar based on the luminance distribution of each pixel in the image;
A third step of specifying the region as a reinforcing bar region when the luminance range in each region sandwiched between the extracted edges includes the calculated standard luminance value;
Reinforcing bar region extraction program characterized by including.