Adhesive application detection device
The coating determination device addresses the challenge of determining adhesive application on building materials by using image processing and machine learning to classify and count adhesive tubes, ensuring appropriate application and preventing flooring material lifting.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- DAIWA HOUSE INDUSTRY CO LTD
- Filing Date
- 2022-08-08
- Publication Date
- 2026-07-02
AI Technical Summary
Existing technologies struggle to determine whether an appropriate amount of adhesive is applied to building materials, particularly when constructors directly apply adhesive to flooring materials.
A coating determination device that includes an image extraction unit, adhesive classification unit, number identification unit, and application amount determination unit, utilizing image processing and machine learning to classify and count adhesive application along the long and short sides of rectangular building materials, projecting images onto a front view for accurate classification.
Enables easy determination of appropriate adhesive application by counting adhesive tubes along the long and short sides, ensuring accurate and efficient application without lifting of flooring materials.
Smart Images
Figure 0007883908000001 
Figure 0007883908000002 
Figure 0007883908000003
Abstract
Description
Technical Field
[0001] The present invention relates to a coating determination device for determining whether the coating amount of an adhesive applied along the long side and short side directions of a rectangular building material is an appropriate amount.
Background Art
[0002] Conventionally, a determination device for determining the coating state of an adhesive applied to a building material has been proposed. For example, the determination device described in Patent Document 1 attaches a coating nozzle and an imaging device to a coating head, and while imaging the adhesive applied from the coating nozzle with the imaging device, determines whether the adhesive is applied at an appropriate position.
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] However, different from Patent Document 1, when applying an adhesive to a building material such as a flooring material, the constructor directly applies the adhesive, and it is difficult to determine whether an appropriate amount of the adhesive is applied.
[0005] The present invention has been made in view of such points, and an object thereof is to provide a coating determination device for an adhesive that can easily determine whether the coating amount of the adhesive applied to a building material is an appropriate amount.
Means for Solving the Problems
[0006] In view of the above-mentioned problems, the adhesive application determination device according to the present invention is an application determination device that determines whether the amount of adhesive applied to a rectangular building material with a long side and a short side along the direction of the long side and the direction of the short side is an appropriate amount, and is characterized by comprising at least: an image extraction unit that extracts images of the building material and images of each adhesive from an overall image including the building material captured by an imaging device, and extracts the outline of the building material from the extracted images of the building material; an adhesive classification unit that classifies the images of the adhesive into images of adhesive in the direction of the long side and images of adhesive in the direction of the short side; a number identification unit that identifies the number of adhesives in the direction of the long side and the number of adhesives in the direction of the short side; and an application amount determination unit that determines whether the amount of adhesive applied to the building material is appropriate based on the number of adhesives in the direction of the long side and the number of adhesives in the direction of the short side.
[0007] According to the present invention, the image extraction unit extracts images of building materials and each adhesive from an overall image including building materials captured by an imaging device, and extracts the contour lines of the building materials from the extracted images of building materials. Examples of this extraction include extraction using machine learning to extract images of building materials, and extraction using image processing such as edge detection.
[0008] The adhesive classification unit classifies adhesive images into images of adhesives along the long side and images of adhesives along the short side. Since the contour of building materials is formed by the short and long sides of the rectangular building material, images of adhesives along the long side and images of adhesives along the short side can be easily classified. Such classification may be performed, for example, by setting straight lines that pass through the pixels constituting the long and short sides, and grouping the pixels of adhesive images that are within a predetermined distance from these lines, or by using cluster analysis as described later.
[0009] By performing this classification, the number identification unit can identify the number of adhesive tubes in the direction of the long side and the number of adhesive tubes in the direction of the short side. Here, the adhesive is stored in a specific tube-type container and is applied by being squeezed out from the tip of this container. Therefore, if the direction in which the adhesive is applied and the number of tubes are known, the amount of adhesive applied can be roughly determined. Accordingly, in this invention, the application amount determination unit can determine whether the amount of adhesive applied to the building material is appropriate based on the number of adhesive tubes in the direction of the long side and the number of adhesive tubes in the direction of the short side. In this way, by imaging the building material with an imaging device, it is possible to easily determine whether the amount of adhesive applied to the building material is appropriate or not.
[0010] Here, as described above, if the images of adhesives can be classified into images of adhesives in the direction of the long side and images of adhesives in the direction of the short side, the classification method by the adhesive classification unit is not particularly limited. However, in a more preferred embodiment, the adhesive classification unit classifies the images of adhesives in the direction of the long side and images of adhesives in the direction of the short side by cluster analysis using the coordinates of the pixels that constitute the images of adhesives.
[0011] According to this embodiment, cluster analysis using the coordinates of pixels that make up the image of the adhesive allows for grouping even areas where the adhesive is applied intermittently into a single group, thus enabling more accurate classification of the image of the adhesive along the long and short sides.
[0012] Here, there is no particular problem if the overall image is captured from a position directly facing the building material, but if it is captured from an angle that is inclined relative to the building material, for example, there is a risk that the number of adhesive strands cannot be accurately calculated. From this point of view, the coating determination device further comprises a projection transformation unit that projects the images of the building material and the adhesive applied to the building material onto a front view image based on the contour lines, and the adhesive classification unit classifies the images of the adhesive into images of adhesive in the direction of the long side and images of adhesive in the direction of the short side.
[0013] According to this embodiment, the projection transformation unit projects the images of the building material and the adhesive applied to the building material onto a front view image based on the contour lines, so that the adhesive classification unit can accurately classify the images of the adhesive into images of the adhesive in the direction of the long side and images of the adhesive in the direction of the short side. [Effects of the Invention]
[0014] According to the present invention, it is possible to easily determine whether the amount of adhesive applied to a building material is appropriate or not. [Brief explanation of the drawing]
[0015] [Figure 1] This is a schematic perspective view illustrating flooring materials to which the adhesive used for determination by the coating determination device of this embodiment has been applied. [Figure 2] This is a schematic diagram of a system including a coating determination device according to this embodiment. [Figure 3] Figure 1 is a control block diagram of the calculation unit of the coating determination device shown. [Figure 4] Figure 2 shows images of the flooring material and adhesive, as well as a schematic diagram to explain the extraction of the flooring material's outline, etc. [Figure 5] Figure 3 is a schematic diagram illustrating the projection transformation section. [Figure 6] Figure 2 is a schematic diagram illustrating the classification of adhesive images by the adhesive classification section. [Figure 7] Figure 1 is a workflow diagram using the coating determination device shown. [Modes for carrying out the invention]
[0016] The coating determination device 10 according to this embodiment will be described below with reference to Figures 1 to 7.
[0017] 1. Regarding building materials 5 and adhesives 4 In this embodiment, as shown in FIG. 1, the coating determination device 10 is a device that determines whether the coating amount of the adhesive 4 applied along the direction X of the long side 51 and the direction Y of the short side 52 on the rectangular floor material (building material) 5 with the long side 51 and the short side 52 is an appropriate amount.
[0018] In this embodiment, the floor material 5 is exemplified as the building material. However, for example, not only interior materials such as gypsum boards, wall materials, and ceiling materials, but also exterior wall materials, slate materials, eaves materials (decorative materials under the eaves), etc. may be used. As the adhesive, for example, a vinyl acetate adhesive, a urethane-based adhesive, etc. may be used, or a rubber-based, acrylic-based, or silicone-based adhesive or these adhesives may be used.
[0019] As shown in FIG. 1, the floor material 5 is a rectangular plate material composed of the long side 51 and the short side 52 in plan view. Among the side surfaces of the floor material 5, male joints 55A and female joints 55B are formed on the side surface in the long side direction X, and male joints 56A and female joints 56B are formed on the side surface in the short side direction Y.
[0020] When laying a plurality of floor materials 5, 5,... after applying the adhesive A described later to the back surface 53 of the floor material 5, the floor material 5 is placed on the floor under the floor, and the male joint 55A (56A) of the floor material 5 is actually joined to the female joint 55B (56B) of the adjacent floor material.
[0021] Here, in order to prevent the laid floor material 5 from lifting from the floor under the floor, it is important to apply an appropriate amount of adhesive to the back surface 53 of the floor material 5. The adhesive 4 is stored in a specific tube-type container and is applied to the back surface 53 of the floor material 5 by being extruded from the tip of this container.
[0022] Specifically, one tube of adhesive 4(41) is applied along the edge of the male tongue 55A (or female tongue 55B) in the long side direction X, and multiple tubes of adhesive 4(42) are applied at equal intervals along the male tongue 56A (or female tongue 56B) in the short side direction Y. Therefore, if the number of tubes of adhesive 4 applied in the direction of application (long side direction and short side direction) can be determined, the amount of adhesive applied can be roughly estimated, and thus it is possible to determine whether the amount of adhesive 4 applied to the flooring material 5 is appropriate.
[0023] From this perspective, the coating determination device 10 according to this embodiment determines whether the amount of adhesive 4 applied to the floor material (building material) 5 along the long side direction X and the short side direction Y is appropriate.
[0024] 2. Hardware configuration of the coating determination device 10 The coating determination device 10 comprises a storage unit 10A, which is composed of ROM, RAM, etc., and stores specifications such as the dimensions of the flooring material 5 and a determination program for determining the appropriateness of adhesive application, and a calculation unit 10B that executes the determination program.
[0025] The coating determination device 10 is connected to an input device 31 and an output device 32. In this embodiment, the input device 31 and the output device 32 may be integrated into a single touch panel display. Data such as the specifications of the flooring material 5 and the determination program are input using the input device 31. In this embodiment, image data captured by the imaging device 20 is input to the input device 31. The data input by the input device 31 is stored in the storage unit 10A. The output device 32 displays the image data captured by the imaging device 20, the calculation results calculated by the calculation unit 10B, etc.
[0026] In this embodiment, the coating determination device 10 consisted of a storage unit 10A and a calculation unit 10B, but it may also include, for example, an input device 31 and an output device 32. In addition to the input device 31 and the output device 32, the coating determination device 10 may further include an imaging device 20, and these may be integrated into a mobile terminal such as a smartphone or tablet.
[0027] 3. Software configuration of the coating determination device 10 In this embodiment, as shown in Figure 3, the coating determination device 10 includes at least an image extraction unit 11, a projection transformation unit 12, an adhesive classification unit 13, a number identification unit 14, an area ratio / spacing calculation unit 15, and a coating amount determination unit 16.
[0028] 3-1. About the image extraction unit 11 As shown in Figure 4, the image extraction unit 11 extracts images of the flooring material 5 and each adhesive 4 from the overall image G1 including the flooring material 5 captured by the imaging device 20, and extracts the contour lines Pa, Pb, Pc, and Pd of the flooring material 5 from the extracted images of the flooring material 5.
[0029] The image extraction unit 11 may, for example, use a support vector machine (SVM) to learn the features of the flooring material 5 from the captured image, using the image of the flooring material 5 captured by the imaging device 20 and the shape features of the flooring material 5 in this image (for example, multiple points along the edge of the flooring material) as training data. This makes it possible to extract an image of the flooring material from any overall image G1 that includes an image of the flooring material 5. Alternatively, a cascade classifier using features such as Haar-like features may be used to identify the flooring material itself, and then the image of the flooring material 5 may be extracted from the image of the identified area using the machine-learned model described above.
[0030] Using a similar method, images of multiple adhesives 4 are extracted from the overall image G1. For example, in this embodiment, it is sufficient to identify at least the number of adhesives 4 (41, 42) in the long side direction X and the short side direction Y, and adhesives 4 may be identified using a cascade classifier that has been trained on the shape of adhesives 4. Alternatively, the converted front image may be binarized or grayscaled, and images of adhesives 4 may be identified from the number of pixels that have a predetermined difference in brightness from the surrounding pixels.
[0031] Furthermore, since the flooring material 5 is rectangular in shape, it may be detected as an outline consisting of the four sides of the flooring material 5 by edge detection (such as edge detection using the Canny method) and generally known straight line detection (such as straight line detection using the Hough method).
[0032] In this case, it is preferable that the image of the back surface 53 of the flooring material 5 is extracted as the image of the flooring material 5. However, for example, the image of the flooring material 5 may also include an image of the side surface of the back surface 53 of the flooring material 5, including the male and female fruits 55A and 55B. Even in such a case, the image of the flooring material 5 including the side surface is larger (has more pixels) than the image of only the back surface 53, so the amount of adhesive 4 applied is determined to be less than the actual amount. Therefore, even in such a case, the application amount determination unit 16 will not make a false determination that the amount of adhesive 4 applied has not reached the appropriate amount.
[0033] 3-2. Regarding the projection transformation unit 12 The projection transformation unit 12 projects the image G2 of the flooring material 5 and the applied adhesive 4 into a front view image G3. The projection transformation method used by the projection transformation unit 12 is not particularly limited, as long as it is possible to project the extracted image G2 of the flooring material 5 into a front view image G3.
[0034] For example, the lengths of the short and long sides of a rectangular flooring material 5 may be input, and it may be determined whether the contour lines Pa~Pd of the corresponding flooring material 5 image G2 correspond to the short or long side. The extracted flooring material 5 image G2 may then be converted to a front view image G3 so that the aspect ratio of the contour lines Pa~Pd matches the ratio of the short and long sides. In this case, the intersection points g1~g4 of the contour lines Pa~Pd are calculated, and the length of the contour lines Pa~Pd is calculated from the coordinates of these intersection points. Note that in the projection transformation shown in Figure 5, pixel dilation and scalding of the contour lines may also be performed, and the method for converting to a front view image G3 is a generally known method, so a detailed explanation is omitted.
[0035] 3-3. Regarding the Adhesive Classification Section 13 The adhesive classification unit 13 classifies the adhesive 4 into images L1 of the adhesive 41 in the long side direction X and images A1 to A5 of the adhesive 42 in the short side direction Y, using cluster analysis with the coordinates of the pixels that make up the image of the adhesive 4, as shown in Figure 6.
[0036] Here, depending on the extraction method in the image extraction unit 11 and the method of adhesive application, a single adhesive 42 may be extracted as multiple intermittent, different adhesives 42. For example, as shown in Figure 5, adhesives 42a and 42b (42c, 42d) may not be extracted as separate lines. In such cases, the number of adhesives 42 is determined for each adhesive 42a and 42b (42c, 42d) by the number identification unit 14, which will be described later, resulting in a higher number being identified than the actual number. As a result, the application amount determination unit 16, which will be described later, may incorrectly determine that the amount of adhesive applied is appropriate. From this perspective, the adhesive classification unit 13 groups (classifies) the adhesives 4 at least before identifying the number of adhesives 4, in accordance with the following.
[0037] Specifically, the adhesive classification unit 13 performs cluster analysis using the coordinates of the pixels constituting the adhesives 41 and 42 extracted by the image extraction unit 11 (the XY orthogonal coordinates of each pixel relative to the overall image G1, described later), and groups (classifies) them into images L1 of adhesive 41 in the long side direction X and images A1 to A5 of adhesive 41 in the short side direction Y. In this embodiment, the cluster analysis results in grouping (classification) adhesives 42a and 42b shown in Figure 5 as one adhesive 42, and adhesives 42c and 42d as one adhesive 42.
[0038] More specifically, the adhesive classification unit 13 performs cluster analysis on the coordinate values of each pixel constituting the adhesive 4 using a Bayesian Gaussian mixture model or the like, and classifies the adhesive 4 into images L1 of adhesive 41 in the long side direction X and images A1 to A5 of adhesive 41 in the short side direction Y.
[0039] Here, the adhesive classification unit 13 estimates the number of clusters, including the distribution of coordinate values of each pixel of the extracted adhesives 41 and 42. Here, only adhesives 41 whose length in the long side direction X (the difference between the minimum and maximum X coordinates of the pixels at both ends) is 90% or more of the length in the long side direction X of the flooring material 5 are extracted. Similarly, only adhesives 42 whose length in the short side direction Y (the difference between the minimum and maximum Y coordinates of the pixels at both ends) is 90% or more of the length in the short side direction Y of the flooring material 5 are extracted. In this way, the classified adhesives 41 and 42 in the long side direction X and short side direction Y can be reliably extracted.
[0040] 3-4. Regarding the number identification unit 14 The number identification unit 14 identifies the number of adhesives 41 in the long side direction L and the number of adhesives 42 in the short side direction Y, as classified by the adhesive classification unit 13. In this embodiment shown in Figure 6, it can identify that there is 1 adhesive 41 in the long side direction L and 5 adhesives 42 in the short side direction Y.
[0041] 3-5. Area ratio / interval calculation unit 15 The area ratio / spacing calculation unit 15 calculates the ratio of the area of the adhesive 4 to the area of the back surface 53 of the flooring material 5 (area ratio of the adhesive 4), and the spacing between adhesives 41 in the long side direction X and between adhesives 42 in the short side direction. As shown in Figure 6, there is only one adhesive 41 in the long side direction X, so only the spacing between adhesives 42 in the short side direction Y is calculated here.
[0042] The area ratio of adhesive 4 can be calculated by dividing the number of pixels in the image of flooring material 5 enclosed by the contour lines Pa to Pd of flooring material 5 by the total number of pixels in the image of all adhesive 4. When calculating the spacing between adhesives 42, the equation of a straight line through which each adhesive 42 passes can be found using the least squares method, and the distance along the long side direction X between the straight lines of the equations of adjacent adhesives 42 is calculated. Alternatively, principal component analysis may be performed on the coordinates of the pixels that make up the image of each adhesive 42 to calculate a straight line (principal component axis) passing through the center of the adhesive 42 in the direction of the first principal component. In addition, the center coordinates of adhesives 42 (images A1 to A5) can be calculated, and the distance to adjacent adhesives can be calculated from the center coordinates.
[0043] Furthermore, when making a determination using the application amount determination unit 16 described later, if the area ratio of the adhesive and the spacing between adhesives are not used as conditions for determining the appropriate application amount or appropriate application method of the adhesive 4, then the area ratio / spacing calculation unit 15 does not need to be provided.
[0044] 3-6. Regarding the coating amount determination unit 16 The application amount determination unit 16 determines whether the amount of adhesive 4 applied to the flooring material 5 is appropriate based on the number of adhesives 41 in the long side direction X and the number of adhesives 42 in the short side direction Y. Specifically, if the number of adhesives 41 in the long side direction X and the number of adhesives 42 in the short side direction Y are equal to or greater than a preset number, it is determined that the amount of adhesive 4 applied is appropriate (the correct amount). In this embodiment, the determination conditions are set so that the number of adhesives 41 in the long side direction X is 1 or more, and the number of adhesives 42 in the short side direction Y is 5 or more, so it is determined that the amount of adhesive 4 applied to the flooring material 5 is appropriate.
[0045] In addition to the above-mentioned determination conditions, the application amount determination unit 16 may also determine that the amount of adhesive 4 applied to the flooring material 5 is appropriate if the area ratio of the adhesive 4 is equal to or greater than a preset value. Furthermore, the application amount determination unit 16 may determine that the application form of the adhesive 4 applied to the flooring material 5 is appropriate if the spacing between adhesives 41 in the long side direction X and between adhesives 42 in the short side direction Y is equal to or less than a preset value.
[0046] The following describes the workflow using the coating determination device 10 with reference to Figure 7. First, in step S1, adhesive 4 is applied to the back surface 53 of the flooring material 5. In step S2, the imaging device 20 is used to photograph the area including the flooring material 5 as the inspection target area, and an overall image G1 is obtained.
[0047] Next, in step S3, the image extraction unit 11 extracts images of the flooring material 5 and the adhesive 4 from the overall image G1 including the flooring material 5 captured by the imaging device 20, and extracts the contour lines Pa to Pd of the flooring material 5 from the extracted images of the flooring material 5.
[0048] Next, in step S4, the projection transformation unit 12 projects the image G2 of the flooring material 5 and the applied adhesive 4 into a front view image G3. In step S5, the adhesive classification unit 13 classifies the image of the adhesive 4 into an image L1 of the adhesive 41 in the long side direction X and images A1 to A5 of the adhesive 42 in the short side direction Y, using cluster analysis based on the coordinates of the pixels that make up the image of the adhesive 4. Because the projection transformation unit 12 projects the image of the flooring material 5 and the adhesive 4 applied to the flooring material 5 into a front view image based on the contour lines Pa to Pd, the adhesive classification unit 13 can accurately classify the image of the adhesive 4 into an image of the adhesive 41 in the long side direction X and an image of the adhesive 42 in the short side direction Y.
[0049] In step S6, the number identification unit 14 counts the number of adhesives 41 in the long side direction X and the number of adhesives 42 in the short side direction Y, as classified by the adhesive classification unit 13. Next, in step S7, the area ratio / spacing calculation unit 15 calculates the area ratio of the adhesives 4 and the spacing between the adhesives 42 in the short side direction.
[0050] In step S8, the application amount determination unit 16 determines whether the amount of adhesive 4 applied to the flooring material 5 is appropriate based on the number of adhesives 41 in the long side direction X and the number of adhesives 42 in the short side direction Y. At this time, the application amount determination unit 16 may also determine whether the amount of adhesive 4 is appropriate and whether the application form of the adhesive 4 is appropriate based on the area ratio of the adhesive 4 and the spacing between the adhesives 42 in the short side direction.
[0051] According to this embodiment, the coating amount determination unit 16 can determine whether the amount of adhesive 4 applied to the flooring material 5 is appropriate based on the number of adhesives 41 in the long side direction X and the number of adhesives 42 in the short side direction Y. Therefore, by imaging the flooring material 5 with the imaging device 20, it is possible to easily determine whether the amount of adhesive 4 applied to the flooring material 5 is appropriate. In particular, by cluster analysis using the coordinates of the pixels that make up the image of the adhesive 4, even areas where the adhesive 4 is applied intermittently can be grouped into a single group, so that the image of the adhesive 4 along the long side direction X and the short side direction Y can be classified more accurately.
[0052] If the amount of adhesive 4 applied is determined to be appropriate, in step S9, the applied adhesive 4 is spread onto the back surface 53 of the flooring material 5. However, the work in step S9 is not required. In step S10, the flooring material 5 on which the adhesive 4 has been spread is laid on the subfloor or the underlayment of the flooring material, and the flooring material 5 is joined together. In this way, the flooring material 5 with the appropriate amount of adhesive 4 applied is laid, so lifting of the flooring material 5 can be avoided.
[0053] Although embodiments of the present invention have been described in detail above, the present invention is not limited to the embodiments described above, and various design modifications can be made without departing from the spirit of the invention as described in the claims. [Explanation of Symbols]
[0054] 5: Flooring material (building material), 4: Adhesive, 10: Coating detection device, 11: Image extraction unit, 12: Projection transformation unit, 13: Adhesive classification unit, 14: Number identification unit, 16: Coating amount determination unit, 20: Imaging device, G1: Overall image, G2: Image of flooring material (building material), G3: Front view image
Claims
1. An application determination device for determining whether the amount of adhesive applied to a rectangular building material with long and short sides, along the direction of the long side and the direction of the short side, is appropriate, An image extraction unit extracts images of the building materials and each adhesive from an overall image including the building materials captured by an imaging device, and extracts the outline of the building materials from the extracted images of the building materials. An adhesive classification unit that classifies the image of the adhesive into an image of the adhesive in the direction of the long side and an image of the adhesive in the direction of the short side. A number identification unit that identifies the number of adhesives in the direction of the longer side and the number of adhesives in the direction of the shorter side, An application amount determination unit determines whether the amount of adhesive applied to the building material is appropriate based on the number of adhesives in the direction of the longer side and the number of adhesives in the direction of the shorter side, An adhesive application determination device characterized by comprising at least the following:
2. The adhesive application determination device according to claim 1, characterized in that the adhesive classification unit classifies the adhesive image into an image of the adhesive in the direction of the long side and an image of the adhesive in the direction of the short side by cluster analysis using the coordinates of the pixels constituting the image of the adhesive.
3. The coating determination device is The system further includes a projection transformation unit that projects images of the building material and the adhesive applied to the building material onto a front view image based on the aforementioned contour lines, The adhesive application determination device according to claim 1 or 2, characterized in that the adhesive classification unit classifies the image of the adhesive into an image of the adhesive in the direction of the long side and an image of the adhesive in the direction of the short side.