JP2019095821A - Stone wall management system, stone wall management method, and computer program - Google Patents

Abstract

To provide a stone wall management system, a stone wall management method and a computer program capable of managing stone materials constituting stone walls respectively without the necessity of attaching an IC tag or the like to the stone materials individually in advance.SOLUTION: A stone wall management system 1 for managing a plurality of stone materials of stone walls formed by laminating the stone materials, comprises a stone wall image data acquisition unit 10 for acquiring stone wall image data obtained by capturing a lamination state, a stone wall elevation view creation unit 22 for creating stone wall elevation view data by extracting a shape of the stone materials through division processing of an image area from the acquired stone wall image data, a stone wall image data extraction unit 25 for extracting stone material image data which indicates the respective stone materials, a stone material management data generation unit 26 for assigning an identification number which allows identification for each extracted stone material, calculating positional information of a corresponding stone material and generating stone material management data in which the identification number and the positional information are associated with each other, and a management data storage unit 40 for storing the generated stone material management data.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an Ishigaki management system, an Ishigaki management method, and a computer program.

  Conventionally, in order to repair stone fences such as fenders and castles, a method of creating a surface profile view has been disclosed in which the surface profile of the stone wall is measured by laser measurement to create a surface profile view (Ishigaki elevation view) of the stone wall (eg, Patent Document 1). In addition, at the time of dismantling and restoration of the stone wall, an IC tag storing the identification number is attached to each stone material that constitutes the stone wall, and image information for each stone material, position information before dismantling, temporary placement position of stone material, stone material It is disclosed that the method of managing the dismantled stone of stone wall can be rebuilt to its original condition before dismantling by storing the result of reusability judgment due to the deterioration of weather and progress of weathering for each identification number. (See, for example, Patent Document 2).

JP 2004-212107 A JP, 2011-065349, A

  However, in the surface layer sectional view of the stone wall created using laser measurement disclosed in Patent Document 1, although it is possible to manage the distribution and the like of the stone material on the surface of the stone wall, individual information and position information about each stone material Etc. can not be managed. In the method disclosed in Patent Document 2, although individual information and position information on individual stone materials can be managed, position information and image information before dismantling of stone materials can be manually used in a personal computer or the like. There is a problem that the work is complicated because it is necessary to register. Furthermore, IC tags etc. must be pasted on individual stone materials beforehand, and in the case of dismantling and restoration work, pasting work such as IC tags can be done at the time of dismantling. In the case of a fall, there is also a problem that management of the fallen stone can not be performed because the pasting work can not be performed in advance.

  The present invention has been made in view of such circumstances, and the stone wall is configured without manually registering the image information and position information of individual stone materials and attaching an IC tag to the stone materials in advance manually. It is an object of the present invention to provide an stone wall management system, an stone wall management method, and a computer program capable of managing image information, position information, and other attribute information of individual stone materials.

  An stone wall management system according to a first aspect of the present invention for achieving the above object is an stone wall management system for managing stone materials of stone walls formed by laminating a plurality of stone materials, an image for acquiring stone wall image data obtained by imaging the stone walls. A data acquisition unit, and an Ishigaki elevation view creation unit that creates the stone wall elevation elevation data of the stone wall showing the laminated state of each stone from the acquired stone wall image data, and a stone area of the created stone wall elevation data For each, by dividing the stone wall image data, a stone image data extraction unit for extracting stone image data indicating each stone and an identification number capable of being identified are given to each of the stone, Stone management data to generate stone management data that includes the position information of the stone, the image data of the stone, and other attribute information for each identification number. Characterized in that it comprises a data generating unit, and a management data storage unit for storing the generated stone management data.

  An Ishigaki management system according to a second invention is characterized in that the image data acquisition unit in the first invention acquires three-dimensional image data including a three-dimensional shape of the surface of the stone wall.

  Further, in the stone wall management system according to the third aspect of the present invention, in any one of the first to second aspects, the Ishigaki elevation view creation unit is configured to divide the stone wall image data into a plurality of regions It is divided into three, and it is characterized by creating Ishigaki elevation data representing the layered situation of each stone material.

  In the stone wall management system according to the fourth aspect of the present invention, in any one of the first to third aspects, the stone material management data generation unit further determines, according to the stone wall elevation data, any of the stone figure regions It is characterized in that the point of is taken as a representative point, and the position information of the stone concerned is used.

  In the stone wall management system according to the fifth aspect of the present invention, in any one of the first to fourth aspects, when a stone falls from the stone wall, the crushed stone image data obtained by imaging the fallen stone is acquired, and the fall occurs. A crushed stone image data acquisition unit which records after adding a crushed stone management number which is an identification number capable of identifying each of the treated stones, an acquired crushed stone image data and the stone image data, and an image analysis method The collapsing stone collating section that determines which stone in the stone management data matches with the collated stone, and acquires position information and other attribute information in the stone wall of the stone from the stone management data. And the management data storage unit stores the collated result in the collapsible stone collating unit as collapsible stone management data.

  Further, according to a sixth aspect of the present invention, there is provided an stone wall management method capable of being executed by a computer that manages each stone material of the stone wall formed by laminating a plurality of stone materials, wherein the computer images the stone wall A process of acquiring stone wall image data, and a process of generating stone wall elevation data of the stone wall indicating the laminated state of each stone material from the acquired stone wall image data, and a region of stone material of the generated stone wall elevation data By dividing the stone wall image data, extracting stone image data representing each stone, and giving an identification number that can be identified to each stone, for each identification number, A process of generating stone management data associated with the position information of the stone, the image data of the stone, and other attribute information, and the generated stone pipe Characterized in that it comprises a step of storing the data.

  A computer program according to a seventh aspect of the present invention is a computer program that can be executed by a computer that manages each stone material of an stone wall formed by laminating a plurality of stone materials, wherein the computer is an Ishigaki image obtained by imaging the stone wall Ishigaki image data acquisition means to acquire data, Ishigaki elevation view creation means to create Ishigaki elevation view data of the stone wall showing the laminated state of each stone material from the acquired Ishigaki image data, Ishigaki elevation view created A stone material image data extraction means for extracting stone material image data indicating each stone material by dividing the stone wall image data for each of the stone material areas of the data, and an identification number that can be identified for each of the stone materials Position information of the stone, image data of the stone, and other attributes for each identification number Characterized in that to function as masonry management data storage means for storing stone management data generating means for generating a stone management data associated, including broadcast, and the generated stone management data.

  According to the present invention, the registration of the image information and position information of the individual stone materials constituting the stone wall, and the manual application of the IC tag etc. to the stone material in advance, without regard to the individual stone materials constituting the stone wall Information, in particular, image information and position information can be managed, and it becomes possible to restore the original state with high accuracy at the time of restoration when the stone wall collapses due to a disaster or the like.

It is a functional block diagram showing typically the composition of the stone fence management system concerning an embodiment of the invention. It is a block diagram which shows typically the structure at the time of using CPU of the computer processing part of the stone wallage management system which concerns on embodiment of this invention. It is a flowchart which shows the flow of the stone fence management processing of the computer processing part of the stone fence management system which concerns on embodiment of this invention. It is a flowchart which shows the flow of the Ishigaki elevation creation processing of the computer processing part of the Ishigaki management system which concerns on embodiment of this invention. It is a flowchart which shows the flow of the multi-scale division | segmentation process of the computer processing part of the stone fence management system which concerns on embodiment of this invention. It is explanatory drawing which shows the merging process of two adjacent area | regions in the multi-scale division process of the computer processing part of the stone fence management system which concerns on embodiment of this invention. It is a flowchart which shows the flow of the stone management data generation process of the computer processing part of the stone fence management system which concerns on embodiment of this invention. It is explanatory drawing which shows the data structure of stone DB of the stone-gage management system which concerns on embodiment of this invention. It is explanatory drawing at the time of superposing the outline of a stone and position information on the image data of the surface layer surface of the stone wall of the stone fence management system which concerns on embodiment of this invention, and displaying it. It is a flow chart which shows a flow of collapsing stone material collation processing of a computer processing part of a stone wall control system concerning an embodiment of the invention. It is a flowchart which shows the flow of the stone material image data collation process of the computer processing part of the stone wallage management system which concerns on embodiment of this invention. It is explanatory drawing which shows the data structure of collapsing stone material collation result DB of the stone wall management system which concerns on embodiment of this invention.

  Hereinafter, the stone wall management system which manages each stone material of the stone wall which laminates | stacks a several stone material based on embodiment of this invention is concretely demonstrated based on drawing. The following embodiments do not limit the invention described in the claims, and all combinations of characteristic items described in the embodiments are essential to the solution. Needless to say, it is not limited.

  In addition, the present invention can be implemented in many different modes, and should not be construed as being limited to the description of the embodiments. The same symbols are given to the same elements throughout the embodiment.

  In the following embodiments, a system in which a computer program is introduced into a computer system will be described, but as is apparent to those skilled in the art, the present invention is embodied as a computer program that can be partially executed by a computer. be able to. Therefore, according to the present invention, an embodiment as a hardware, an embodiment as software, or a combination of software and hardware as an stone wall management system for managing each stone material of stone walls formed by laminating a plurality of stones is described. An embodiment can be taken. The computer program can be recorded on any computer readable recording medium such as a hard disk, a DVD, a CD, an optical storage device, and a magnetic storage device.

  According to the embodiment of the present invention, information on each stone constituting the stone wall, in particular, shape information and position information is managed without attaching an IC tag or the like to each stone constituting the stone wall in advance. It is possible to restore the original state with high accuracy at the time of stone wall restoration etc.

  FIG. 1 is a functional block diagram schematically showing the configuration of the stone wall management system according to the embodiment of the present invention. The stone fence management system 1 according to the embodiment of the present invention comprises at least an image data acquisition unit 10, a computer processing unit 20, an input / output unit 30, a management data storage unit 40, and a fallen stone material image data acquisition unit 50.

  The image data acquisition unit 10 acquires image data of the surface layer (the layered state of the stone material) of the stone wall to be managed, is configured of an optical digital camera or the like, and images the surface layer to obtain an image of the surface layer. The imaging unit 11 for acquiring data, and the laser measurement unit 12 for measuring the position information of the working reference point disposed on the surface layer of the stone wall by irradiating the laser beam or the like.

  The computer processing unit 20 is configured by hardware including a CPU such as a computer and an arithmetic processing unit such as an MPU, and the image data of the surface of the stone wall acquired by the image data acquisition unit 10 and the position of the work reference point From the information, Ishigaki image data processing unit 21 that generates Ishigaki image data by orthographic projection conversion of the image data of the surface layer of Ishigaki, From the generated Ishigaki image data, Ishigaki elevation view data of Ishigaki showing the laminated state of each stone Each of Ishigaki Elevation Drawing Unit 22 that generates the image and stone is given an identification number that can be identified, and the position information of the stone, image data, and other attribute information are included for each identification number. A stone management data generation unit 26 is provided which generates stone management data associated with each other.

  FIG. 2 is a block diagram schematically showing a configuration in the case where the CPU of the computer processing unit 20 of the Ishigaki management system according to the embodiment of the present invention is used. The computer processing unit 20 according to the embodiment of the present invention includes at least a CPU (central processing unit) 201, a memory 202, a storage device 203, an I / O interface 204, an imaging interface 205, a portable external storage device 206, and a communication interface 207. And an internal bus 208 for connecting the hardware described above.

  The CPU 201 is connected to the above-described hardware units of the computer processing unit 20 via the internal bus 208, controls the operations of the above-described hardware units, and according to the computer program 200 stored in the storage device 203. , Perform various software functions. The memory 202 is composed of volatile memory such as SRAM, SDRAM, etc. The load module is expanded when the computer program 200 is executed, and stores temporary data and the like generated when the computer program 200 is executed.

  The storage device 203 is configured by a built-in fixed storage device (such as a hard disk or an SSD), a ROM, or the like. The computer program 200 stored in the storage device 203 is downloaded by the portable disk drive 206 from a portable recording medium 210 such as a DVD or a CD-ROM in which information such as a program and data is recorded. It is expanded to the memory 202 and executed. Of course, it may be a computer program downloaded from an external computer connected via the communication interface 207.

  The storage device 203 may include a management data storage unit 40 described later. However, in the present embodiment, the stone management data having a large file size is configured on the external computer or the external disk.

  The communication interface 207 is connected to the internal bus 208, and by connecting to an external network such as the Internet, LAN, WAN, USB, etc., it is possible to transmit and receive data with an external computer, an external disk, etc. .

  The I / O interface 204 is connected to the input / output unit 30 such as the display 31 and the keyboard 32, receives data input, and displays a predetermined image on the display 31 such as a CRT display or a liquid crystal display. The imaging interface 205 is connected to the image data acquisition unit 10 and the crushed stone image data acquisition unit 50, and acquires predetermined image data.

  Returning to FIG. 1, the Ishigaki elevation view creation unit 22 divides the stone wall image data into individual stone shapes by an image area division process of dividing it into a plurality of areas (also referred to as “objects”) by an area analysis method by image analysis. A stone area dividing unit 23 for recognizing an area, a polygon data generating unit 24 for generating Ishigaki elevation view data, which is polygon data in vector format indicating the outline of the stone shape from the divided individual stone shape areas, and A stone material image data extraction unit 25 is provided which extracts stone material image data indicating individual stone materials by dividing and extracting the stone wall image data for each region of stone material polygon data of the stone wall elevation data.

  In addition, the stone management data generation unit 26 is an identification number giving unit 27 for giving an identification number that can be identified for each stone in the stone wall elevation data created by the polygon data creation unit 24; A stone management data is generated by correlating stone image data, stone position information, and attribute information of each stone using the position information calculation unit 28 for calculating position information and the stone identification number as keys, and management data A matching unit 29 stored in the storage unit 40 is provided. In the present embodiment, the position information of the stone is used as position information determined uniquely for each stone by calculating the barycentric position of the polygon figure extracted from the stone image data as three-dimensional or two-dimensional coordinate information. ing. However, the coordinate information of an arbitrary point in the polygon figure may be used as the position information of the stone, and the coordinate information of the stone polygon data itself may be used as it is as the position information.

  In addition, when the stone wall collapses due to a disaster or the like, the computer processing unit 20 acquires image data (collapsed stone material image data) of the collapsed stone material by the collapsed stone material image data acquisition unit 50 described later and acquired collapsed stone material image data Is stored in the crushed stone DB 42 in the management data storage unit 40.

  The input / output unit 30 is a display 31 for displaying information such as stone wall image data, stone polygon image data, stone polygon data of stone wall, stone DB 41 stored in stone management data, and the like, and stone management data It comprises a keyboard 32 for inputting attribute information and the like of a stone to be a part.

  The management data storage unit 40 is configured of a storage medium such as a hard disk, and has a stone DB 41 for storing stone management data, and a crushed stone DB 42 for storing crushed stone image data and their attribute information.

  The collapsing stone image data acquisition unit 50 acquires collapsing stone image data, which is image data of collapsing stone material, and has the same configuration as the image data acquiring unit 10 described above, and the surface of the collapsing stone material Image capturing unit 51 for capturing image data.

  FIG. 3 is a flow chart showing the flow of the stone fence management process of the computer processing unit 20 of the stone fence management system according to the embodiment of the present invention. In the stone wall management process shown in FIG. 3, the image data of the stone wall is acquired from the image etc. of the surface layer of the stone wall, and the information (stone material management data) of individual stone materials constituting the stone wall is generated and managed. As shown in FIG. 3, the computer processing unit 20 of the Ishigaki management system according to the embodiment of the present invention first uses the image capturing unit 11 of the image data acquisition unit 10 to obtain image data of the stone wall to be managed. The data acquisition process is executed, and the acquired image data is received (step S100).

  The computer processing unit 20 causes the imaging unit 11 of the image data acquisition unit 10 to capture an image of a management target stone wall. Here, in order to acquire the three-dimensional shape of the surface of the stone constituting the stone wall and the stone wall, the imaging unit 11 acquires a plurality of stereo cameras for acquiring the image of the stone wall from at least two different positions or moving in the air. The camera mounted on a flying object (for example, a drone) that acquires an image of Ishigaki from the position of the camera, a single camera, changing at least one of the shooting position and the shooting direction, shooting a plurality of photographs between adjacent areas For example, a method of overlapping and photographing is used. Also, in order to add positional information to the image data, a work reference point is installed at an arbitrary point on the surface layer of the stone wall, and imaging is performed so that the work reference point can be read in the image data. At the same time, the computer processing unit 20 performs reference point surveying based on public coordinates by the laser measurement unit 12 of the image data acquisition unit 10 on the working reference point, and receives the measurement result. The position information in this case uses position information based on public coordinates, but may be position information having a predetermined position as an origin for each stone wall to be managed. The image of the stone wall captured by the imaging unit 11 may be configured to be temporarily stored in a storage medium such as the management data storage unit 40. In addition, the image of Ishigaki may be diverted from what has already been taken.

  The computer processing unit 20 receives the stone wall image (at least two or more images as described above) captured by the image data acquisition unit 10 and received by the stone wall image data processing unit 21 and the position information of the work reference point It is used to generate three-dimensional image data. In this embodiment, the imaging position and imaging direction of the camera are estimated from a plurality of images by SfM (Structure / Shape from Motion) processing, and the imaging position of the camera estimated by MVS (Multi-View Stereo) processing. And reconstruct the three-dimensional shape by stereo matching processing of the image based on the photographing direction and generate three-dimensional image data. This three-dimensional image data is obtained by approximating the surface shape (three-dimensional shape) by a TIN (Triangulated Irregular Network; irregular triangle network) or the like and pasting the texture of the corresponding image on each surface.

  The computer processing unit 20 causes the stone wall image data processing unit 21 to generate stone wall image data, which is ortho image data, from the three-dimensional image data as image data for extracting stone material image data. Here, “ortho image data” is image data obtained by performing orthographic projection conversion of three-dimensional image data of the surface of the stone wall in the horizontal direction from the front of the stone wall. Based on the generated ortho image data, the image data of the surface layer of the stone wall can be made as the image data displayed at the correct position without distortion. The three-dimensional image data and the ortho image data (Ishigaki image data) may also be configured to be temporarily stored in a storage medium such as the management data storage unit 40. Moreover, you may divert what is already produced for ortho image data.

  The computer processing unit 20 executes the Ishigaki elevation creation processing in the Ishigaki elevation creation unit 22 (step S200). FIG. 4 is a flow chart showing the flow of the Ishigaki elevation creation processing of the computer processing unit 20 of the Ishigaki management system according to the embodiment of the present invention.

  As shown in FIG. 4, in the stone material region dividing unit 23 of the Ishigaki elevation view making unit 22, the computer processing unit 20 generates a region by image analysis from the stone wall image data (in the present embodiment, the ortho image data described above). The regions (objects) of the individual stone materials constituting the stone wall are generated by the division method, and the outline thereof is determined (step S210). FIG. 5 shows an example of an area dividing method by image analysis in the stone material area dividing unit 23 of the computer processing unit 20 of the stone wall management system according to the embodiment of the present invention, which is a flow when using a method called multi-scale division processing. Is a flowchart showing

  As shown in FIG. 5, the computer processing unit 20 causes the display 31 to display Ishigaki image data, and adjusts the tone and brightness of the Ishigaki image data based on the input from the keyboard 32 or the like (step S212). The computer processing unit 20 determines the values of setting parameters used for multi-scale division processing and contour tracking processing to be described later based on the adjusted stone wall image data, and accepts setting input as an initial value of the setting parameters (step S214). The size and color of the stone material constituting the stone wall are characterized for each stone wall, and therefore, by setting the setting parameters for each stone wall, it is possible to improve the accuracy of contour determination by the contour tracking process.

  The computer processing unit 20 updates processing parameters for multi-scale division processing (step S216). The computer processing unit 20 starts processing each of the pixels of the image data as a region (one pixel is one region) in the stone material region dividing unit 23, and if two adjacent regions satisfy a predetermined condition, the region is merge. When determination of merging is performed on all the regions of the image data, processing of merging two adjacent regions is performed again, and the feature amount (color variation) is similar by repeatedly performing processing. Areas can be merged to extract areas per stone. Here, a parameter that determines whether to merge two regions is called a scale parameter SP, and the scale parameter SP is added each time the merging process is repeated. The computer processing unit 20 defines the upper limit value of the scale parameter SP, and determines whether the scale parameter SP has reached the upper limit value (step S218).

As processing parameters, in addition to scale parameter SP, spectrum reference F _color weight w _color , shape reference F _shape weight w _shape , compactness reference f _cmpct weight w _cmpct , smoothness reference f _smth weight w _smth , And the specified value of the area of the merged area (the upper limit of the area of the generation area).

  If the computer processing unit 20 determines that the scale parameter SP has not reached the upper limit (step S218: NO), the computer processing unit 20 takes out two adjacent areas from the image data (step S220), and this step It is determined whether or not there are two remaining areas which can be taken out in S220, that is, whether or not all the two areas have been taken out in step S220 (step S222). If the computer processing unit 20 determines that extraction of all two areas is completed (step S222: YES), the computer processing unit 20 returns the process to step S216 and repeats the above-described process.

  If the computer processing unit 20 determines that extraction of all two areas has not been completed yet (step S222: NO), the computer processing unit 20 combines the two adjacent areas extracted in step S220. It is determined whether (merging area) is less than or equal to a specified value (step S224). Note that the unit of the default value (upper limit value of the area of the generation area) of the merging area of two adjacent areas can use the number of pixels or the actual size (when using the actual size, per pixel of the image data). Side length (one pixel resolution, for example, 1 cm / pixel) is specified. Further, as a predetermined value (upper limit value of the area of the generation area) of the merged area of the two adjacent areas, an upper limit value which can be assumed of the size of the stone constituting the stone wall to be managed is set. By setting the upper limit value of the area of the generation area when two adjacent areas are merged in this way, when the stones having similar colors are adjacent to each other, the area of these stones is 1 It can be prevented from being merged as an area of two stones.

  If the computer processing unit 20 determines that the merged area is larger than the specified value (step S224: NO), the computer processing unit 20 returns the process to step S220 and repeats the above-described process. If the computer processing unit 20 determines that the merged area is less than or equal to the specified value (step S224: YES), the computer processing unit 20 calculates and evaluates the evaluation value F of two adjacent areas (step S226). .

  FIG. 6 is an explanatory view showing a merging process of adjacent two areas in the multi-scale division process of the computer processing unit 20 of the Ishigaki management system according to the embodiment of the present invention. As shown in FIG. 6, one of the two adjacent areas is an area p, the other is an area q, and the merged area when the areas p and q are merged is an area r. The evaluation value F indicates a change in heterogeneity before and after merging of two adjacent regions, and is expressed as Expression (1).

In Equation (1), the spectrum reference F _color indicates a change in spectral heterogeneity, and the shape reference F _shape indicates a change in shape heterogeneity. Further, a spectral reference F _color which is a change in the heterogeneity of the spectrum is expressed as Expression (2) using standard deviation values in each band of the pixels constituting the region. The suffixes p and q indicate the two areas to be merged shown in FIG. 7, and r indicates the area after merging. Further, n indicates the number of pixels constituting the area, σ _i and w _i indicate the standard deviation value and its weight in the i-th band, for example, R, G and B bands in the case of a color image, N Indicates the number of bands.

Further, the shape reference F _shape which is a change in shape heterogeneity is expressed as Expression (3) using a compactness reference f _cmpct and a smoothness reference f _smth which define the _shape of the region.

Here, the compactness criterion f _cmpct indicates a change in compactness heterogeneity, and the smoothness criterion f _smth indicates a change in smoothnessness heterogeneity. Then, the compactness criterion f _cmpct is expressed as Expression (4) from the peripheral length of the region and the number of pixels (when the area is equal, generation of a region with a shorter peripheral length is desirable), and smoothness criterion f _smth Is expressed as Equation (5) using the perimeter of the region and the short side of the range of the region (the generation of a region whose perimeter is as short as possible with respect to the range of presence is regarded as ideal) ). Here, n indicates the number of pixels forming the area, l indicates the perimeter of the area, and d indicates the short side length of the bounding box that includes the area. Note that, instead of the short side length of the bounding box including the area, the long side length of the bounding box or the diagonal length of the bounding box can be used for d.

  Referring back to FIG. 5, when the evaluation value F of the two adjacent areas is calculated by the equation (1) by the process described above, the computer processing unit 20 determines whether the calculated evaluation value F is less than or equal to a predetermined value. (Step S228). When the computer processing unit 20 determines that the evaluation value F is larger than the predetermined value (step S228: NO), the computer processing unit 20 returns the process to step S220 and repeats the above-described process.

  If the computer processing unit 20 determines that the evaluation value F is less than or equal to the predetermined value (step S228: YES), the computer processing unit 20 merges two adjacent areas (step S230), and the process proceeds to step S220. return. In the present embodiment, the merging process is applied when the evaluation value F according to Equation (1) does not exceed the square of the scale parameter SP.

  That is, when the evaluation value F is within the predetermined value (square of the scale parameter SP), adjacent two areas are merged, and when the merged area of the two adjacent areas is larger than the predetermined value, the evaluation value F is If it is larger than the default value, adjacent two areas are not merged.

  If the computer processing unit 20 determines that the scale parameter SP has reached the upper limit value (step S218: YES), the computer processing unit 20 executes the contour tracking process (step S232). Specifically, the outline of each of the regions determined by the above-described multi-scale division processing is detected.

  The computer processing unit 20 superimposes the detected contour line on the image data of the management target stone wall and displays it on the display 31, and receives an input of final contour tracking data editing (step S234). Here, the final contour tracking data editing means that there is an error in the contour detected in step S232 (if the contour does not match the shape of the stone wall, or if one stone wall is divided into a plurality of divided regions) In the case where it is detected as an outline, or when a plurality of stone walls are detected as an outline of one area, etc.), the outline displayed on the display 31 is manually added by the keyboard 32 or the like. Or it is the process which receives the input of deletion.

  The computer processing unit 20 determines whether a confirmation input for fixing the result of the multi-scale division processing has been received (step S236). When the computer processing unit 20 determines that the confirmation input is not received (step S 236: NO), the computer processing unit 20 receives the change input of the setting parameter described above from the keyboard 32 or the like (step S 238), It returns to S216 and repeats the processing mentioned above.

  If the computer processing unit 20 determines that the confirmation input has been received (step S 236: YES), the computer processing unit 20 performs the contour smoothing process to make the contour line smoother (step S 240), and multiscale The division process (step S210) is ended, and the process proceeds to step S250 in FIG. In addition, each area | region divided | segmented by the above process turns into the shape of each stone material.

  Referring back to FIG. 4, in the computer processing unit 20, in the polygon data creating unit 24 of the Ishigaki elevation creating unit 22, for each area divided by the stone region dividing unit 23 (shape of each stone forming the stone fence) The stone wall elevation data is created by extracting stone polygon data indicating the outline of the stone (step S250). The Ishigaki elevation data may also be temporarily stored in a storage medium such as the management data storage unit 40 or the like. Also, the elevation data of the Ishigaki used in the present embodiment is not automatically generated by image recognition, but is created by a plotting machine etc., is digitizing from a photograph of Ishigaki, other various manufacturing methods etc. You may divert the created Ishigaki elevation data.

  Returning to FIG. 3, the computer processing unit 20 causes the stone image data extraction unit 25 of the Ishigaki elevation view creation unit 22 to generate stone polygons within the range of each area of the stone polygon data of the created stone wall elevation data By dividing and extracting the image data, stone image data indicating individual stones is extracted (step S300). When extraction of stone material image data is completed for all stone polygon data, the computer processing unit 20 advances the process to step S400.

  The computer processing unit 20 executes, in the stone management data generation unit 26, a stone management data generation process in which stone management data is generated for each of the stones constituting the stone wall to be managed (step S400). FIG. 7 is a flowchart showing a flow of stone management data generation processing of the computer processing unit 20 of the stone wall management system according to the embodiment of the present invention.

  As shown in FIG. 7, in the computer processing unit 20, in the identification number giving unit 27 of the stone material management data generation unit 26, the Ishigaki elevation data created in the polygon data creation unit 24 in the Ishigaki elevation creation unit 22 An identification number capable of uniquely identifying a stone (or stone image data) in the stone wall to be managed is assigned to all the stone polygon data divided into regions (step S410).

  In the position information calculation unit 28 of the stone management data generation unit 26, the computer processing unit 20 calculates the barycentric position of each stone polygon data of the stone wall elevation data as the position information of the stone as three-dimensional or two-dimensional coordinate information. (Step S420) In the associating unit 29 of the stone management data generation unit 26, the identification number and the position information are associated with each stone (step S430). The position information of the stone may be coordinate information of an arbitrary point within the polygon figure, and the coordinate information of the stone polygon data itself may be used as it is as position information.

  Returning to FIG. 3, the computer processing unit 20 stores the generated stone management data in the stone DB 41 of the management data storage unit 40 (step S500). At this time, the stone image data is also stored in the stone DB 41 in association with the identification number.

  FIG. 8 is an explanatory view showing a data structure of a stone material DB of the stone fence management system according to the embodiment of the present invention. As shown in FIG. 8, the stone DB 41 according to the present embodiment stores the stone wall number 41a, the identification number 41b, the other attribute 41c, the position information 41d, and the stone image data 41e as stone management data for one stone. Is configured. Here, the stone wall number 41a is an identification number for identifying the stone wall to be managed, and as the other attribute 41c, it stores, for example, longitudinal longitudinal, lateral longitudinal, weight, stone quality, processing status, arrangement position, etc. It is configured to be possible.

  FIG. 9 is an explanatory view of a case where Ishigaki elevation data and position information are superimposed and displayed on the image data of the surface of the stone wall of the stone wall management system according to the embodiment of the present invention.

  As shown in FIG. 9A, in the entire stone wall, for each stone, stone polygon data, which is an outline of stone image data, and a barycentric position are displayed in an overlapping manner. That is, as can be seen from the partial enlarged view of FIG. 9B, the individual numbers indicate the identification numbers, and the black circles indicate the barycentric positions of the respective stone materials. Since the stone DB 41 is associated with the stone wall elevation view data (stone polygon data which is the outline of stone image data) by the identification number, the stone DB is managed in the place where the identification number is currently displayed. Other values of attribute information may be displayed. In addition, the shape of the stone polygon data of the stone wall elevation data may be color-coded and displayed by ranking expression or individual value classification according to the numerical value of the attribute information and the like.

  Next, the process in the case where the stone wall which manages each stone material by stone material DB41 as stone material management data collapses by disaster etc. is demonstrated. First, the computer processing unit 20 causes the crushed stone image data acquisition unit 50 to acquire the image data of the fallen stone (broken stone image data), and the crushed stone image comparison unit 60 acquires the crushed stone image data into the stone management data of the stone DB 41 Execute the collapsing stone material collating process which collates with the stone image data in the inside. By the collapsing stone material collating process, it becomes possible to extract the information of the stone material in the stone material DB similar to the fallen stone material, and it is possible to acquire the positional information and other attribute information of the fallen stone material.

  First, the fallen stone image data acquisition unit 50 is configured by an optical camera, and photographs the surface of the fallen stone wall stone to obtain the fallen stone image data. After acquiring the crushed stone image data, the identification number (collapsing stone management number) for identifying the fallen rock materials is sequentially numbered and stored in the crushed stone DB 42 together with the crushed stone image data. In addition, in the collapsing stone DB 42, not only collapsing stone image data, but also attribute information such as vertical and horizontal length, weight, stone quality, processing status, etc., the place where the stone has fallen, and temporary storage It is also possible to manage the position information etc. of the temporarily placed place of.

  FIG. 10 is a flow chart showing the flow of collapsing stone material collating processing of the computer processing unit 20 of the stone fence management system according to the embodiment of the present invention. As shown in FIG. 10, the computer processing unit 20 causes the collapsing stone collating unit 60 to read any one collapsing stone image data from the collapsing stone DB 42 (step S600). Then, the stone image data collating process is executed to collate the read fallen stone image data with the stone image data (stone image data stored in the column of stone image data 41 e of FIG. 8) managed by the stone DB 41 (Step S610).

  Specifically, the computer processing unit 20 extracts the feature points of the fallen stone image data and the stone image data for each of the stone management data stored in the stone DB 41, and the similarity degree is extracted based on the extracted feature points. The identification numbers of a predetermined number of stones are extracted in descending order of similarity. Here, the feature points of the fallen stone image data and the stone image data are points having some change from adjacent regions such as corners or edges in the image data. The image data of the stone wall (collapsible stone material image data and stone material image data) to be collated differs in the acquisition state of each image data (shooting equipment (camera), shooting conditions (illumination, shooting angle), image resolution etc.) ing. Therefore, it is necessary to use a feature point extraction operator that is robust against changes in image resolution (scale), rotation, and brightness. Representative feature point extraction operators include Scale-Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF), and Features from Accelerated Segment Test (FAST). Furthermore, when it is difficult to shoot directly from the front, for example, Affine-SIFT (a method of improving SIFT) is used to perform good matching even when the shape of the subject is deformed and captured. It can also be used.

  FIG. 11 is a flow chart showing a flow of stone image data collation processing of the computer processing unit 20 of the stone fence management system according to the embodiment of the present invention. As shown in FIG. 11, the computer processing unit 20 extracts feature points of the crushed stone image data read in step S600 (step S611).

  The computer processing unit 20 reads any one stone image data from the stone image DB 41 (step S612). Also, the computer processing unit 20 extracts feature points of the stone material image data read in step S612 (step S613). Note that for stone image data stored in the stone DB 41, feature points are extracted in advance and stored in the stone DB 41 in association with the identification number, and the stone image is stored together with the stone image data in step S612. You may comprise so that the feature point of data may be read.

  The computer processing unit 20 calculates, for each feature point of the fallen stone image data, a pair (also referred to as a matching point or a corresponding point) of feature points of the stone image data whose feature amount most matches the feature point (step S614). . The computer processing unit 20 evaluates the similarity of the feature points based on the pair of feature points calculated in step S613 (step S615). Assuming that the surface of the stone wall is substantially flat, ideally, the matched points (pairs of feature points) undergo projection transformation of coordinates by plane-to-plane projection transformation. The projective transformation can calculate transform coefficients (projective transform coefficients) by giving four or more corresponding points (each image coordinate of matching points) (if more than 4 points, it is computed by the least squares method) To do).

  The computer processing unit 20 excludes outliers with respect to the feature point pair (matching point) calculated in step S614 (step S616). In general, the matching points include mis-matching (mismatching). Therefore, outliers are excluded by applying a method called RANSAC (Random sample condition) to determine a projective transformation coefficient. Specifically, in order to determine a projective transformation coefficient, M (≧ 4) points are randomly sampled from among the matching points, and thereby the projective transformation coefficient is calculated. Then, with respect to matching points (here, Pi and Qi) which were not used for calculation of the projective conversion coefficient, coordinate conversion of Pi is performed by the calculated projective conversion coefficient (here, a point after coordinate conversion is set as Ri And the number N of matching points whose difference in distance between Qi and Ri is within a predetermined value. Such a trial is performed for a fixed number of times, and the value of N at the time of trial when N becomes maximum (this is assumed to be Nmax) is updated and stored. After that, in the trial when Nmax is reached, final projective transformation coefficients are calculated again using all matching points at which the difference in distance after conversion is within the allowable value. The number of trials may be specified in advance, or may be determined statistically using the number of matching points and the reliability (for example, 95%).

  The computer processing unit 20 determines whether or not the feature points of all the stone image data stored in the stone image DB 41 have been collated (step S617). If the computer processing unit 20 determines that there is stone image data whose feature points have not been collated yet (step S617: NO), the computer processing unit 20 returns the process to step S612 and repeats the above-described process. If the computer processing unit 20 determines that the feature points of all the stone image data have been collated (step S617: YES), the computer processing unit 20 ends the stone image data collating process.

  Returning to FIG. 10, the computer processing unit 20 collates the identification number of the fallen stone (collapsed stone management number) and the collated stone image data with the collation result (identification number of the stone image data having a high degree of similarity). The stone matching result DB 43 is updated (step S620). Specifically, the identification numbers of the predetermined number of pieces of stone material image data are extracted in descending order of Nmax calculated in step S616 (in order of similarity), and the collapsing stone material collation result DB 43 is updated.

  The computer processing unit 20 determines whether all collated stone material image data stored in the collapsible stone image DB 42 have been collated (step S630).

  If the computer processing unit 20 determines that collated stone material image data not yet collated exists (step S630: NO), the computer processing unit 20 returns the process to step S600 and repeats the above-described process. If the computer processing unit 20 determines that all the crushed stone material image data has been collated (step S630: YES), the computer processing unit 20 outputs the collated result (content of collated stone collated result DB 43) to the display 31 etc. (Step S640), the collapsing stone reference process ends.

  In addition, in order to reduce the candidates of stone image data (stone image data stored in stone DB 41) to be collated with the fallen stone image data, only the stone image data corresponding to the vicinity of the fallen position of the fallen stone is collated It can be done. Since the stone DB 41 stores stone image data as well as positional information of the stone in the stone wall, only stone image data in a positional range that can be associated with the fallen stone can be targeted for verification. This makes it possible to reduce the number of stone image data to be collated and to reduce the number of collated errors.

  Further, by performing projective conversion of the image in the above-described stone material image data collating process S610, even if there is a slight inclination in photographing the fallen stone material image data, the image can be transformed (deformed) in substantially the same direction. Also, for stone image data collation processing S610, not based on feature points but, for example, similarity judgment based on a normalized cross correlation coefficient between an image projected and converted by projection conversion and a stone material image data image is added By applying it, it becomes possible to further suppress the image matching error.

  FIG. 12 is an explanatory view showing a data structure of collapsing stone material collation result DB 43 of the stone wall management system according to the embodiment of the present invention. As shown in FIG. 12, the collapsing stone collating result DB 43 extracts the identification numbers of four stones in descending order of the degree of similarity in the collapsing stone management number 43a, collapsing stone image data 43b and similar stone number 43c And the other attribute information 43d. The stone image data similar to the fallen stone image data is not limited to four. In addition, for example, a worker who has confirmed the collation result can store a comment or the like for the collation result from the keyboard 32 or the like in the other attribute information 43d.

  As described above, according to the stone wall management system 1 according to the present embodiment, the surface layer surface of the stone wall is imaged to generate image data (for example, ortho image data), based on the generated image data. Information (attribute information, stone material image data, etc.) of individual stone materials forming the stone wall can be managed by the stone material DB 41 together with the identification number and the position information. Moreover, since the image data (stone image data) of each stone material can be automatically extracted from the image data of Ishigaki, the creation of the Ishigaki medical chart of stone wall can be supported.

  In addition, when the stone wall collapses due to a disaster etc., the position in the stone wall of the fallen stone by collating the image data of the fallen stone (broken stone image data) with the stone management data (stone image data) of the stone DB 41 Since the information can be specified, the restoration can be easily performed at the time of stone wall restoration.

  The present invention is not limited to the above-described embodiment, and various changes, improvements, and the like can be made within the scope of the present invention.

DESCRIPTION OF SYMBOLS 1 Ishigaki management system 10 Image data acquisition part 11 Imaging part 20 Computer processing part 22 Ishigaki elevation view preparation part 26 Stone material management data generation part 30 I / O part 40 Management data storage part 50 Broken stone material image data acquisition part 60 Broken stone material collation part

Claims (7)

In the stone wall management system which manages each stone material of the stone wall which laminates a plurality of stone materials,
Ishigaki image data acquisition unit for acquiring Ishigaki image data obtained by imaging the Ishigaki;
Ishigaki Elevation Creation Department, which creates Ishigaki Elevation View data of the stone walls from the acquired Ishigaki image data, showing the layered condition of each stone,
A stone image data extraction unit that extracts stone image data indicating each stone by dividing the stone wall image data for each of the stone areas of the created Ishigaki elevation data;
Stone management data associated with each stone by assigning an identification number that can be identified, and for each identification number, location information of the stone, image data of the stone, and other attribute information Stone management data generation unit to generate
And a management data storage unit for storing generated stone management data.   The Ishigaki management system according to claim 1, wherein the image data acquisition unit acquires three-dimensional image data including a three-dimensional shape of the surface of the stone wall.   The Ishigaki elevation drawing creation section divides the Ishigaki image data into a plurality of areas by an area analysis method by image analysis, and produces Ishigaki elevation drawing data representing the laminated state of each stone material. The stone wall management system according to any one of claims 1 to 2.   The said management data generation part makes an arbitrary point in the area | region figure of each stone the representative point from the said Ishigaki elevation data, It makes it the positional information on the said stone, It is characterized by the above-mentioned. Ishigaki management system described in any one. When stone falls from the stone wall, collapsing stone image data is obtained by imaging the fallen stone, and recording is given after giving a collapsing stone management number which is an identification number capable of identifying each of the fallen stones. Collapsing stone image data acquisition unit,
By collating the acquired fallen stone image data with the stone image data by an image analysis method, it is determined which stone in the stone management data matches the fallen stone, and the stone management data is determined from the stone management data. It has a collapsing stone collaborator that acquires location information and other attribute information in the stone wall,
The stonewall management system according to any one of claims 1 to 4, wherein the management data storage unit stores, as management data, a collation result in the colliding material collating unit. In the stone wall management method that can be executed by a computer that manages each stone material of the stone wall formed by laminating multiple stone materials,
The computer is
Acquiring Ishigaki image data obtained by imaging the Ishigaki;
Creating an Ishigaki elevation view data of the stone wall indicating the laminated state of each stone material from the acquired stone wall image data;
Extracting stone image data indicating each stone by dividing the stone wall image data for each of the stone regions of the created Ishigaki elevation data;
Stone management data associated with each stone by assigning an identification number that can be identified, and for each identification number, location information of the stone, image data of the stone, and other attribute information Generating the
And a step of storing generated stone management data. In a computer program that can be executed by a computer that manages each stone of an stone wall formed by laminating a plurality of stones,
The computer,
Ishigaki image data acquisition means for acquiring Ishigaki image data obtained by imaging the Ishigaki;
Ishigaki elevation view creation means for creating the Ishigaki elevation view data of the stone wall showing the laminated state of each stone material from the acquired stone wall image data,
Stone image data extraction means for extracting stone image data indicating each stone by dividing the stone wall image data for each of the stone regions of the created Ishigaki elevation data;
Stone management data associated with each stone by assigning an identification number that can be identified, and for each identification number, location information of the stone, image data of the stone, and other attribute information What is claimed is: 1. A computer program comprising: a stone management data generation unit configured to generate and a stone management data storage unit configured to store the generated stone management data.