Blade installation quality inspection method and related equipment

Abstract

The invention provides a blade installation quality inspection method and related equipment. The method comprises the following steps: acquiring an installation scene picture; inputting the installation scene picture into an image recognition model, and performing target recognition by the image recognition model to obtain target recognition information; if the target identification information contains complete target object positioning data, obtaining positioning data of a graduated scale and positioning data of a to-be-measured mark from the target identification information; determining an inclination angle and vertex coordinates of the graduated scale according to the positioning data of the graduated scale; determining a zero graduation line of the calibrated scale based on the inclination angle and vertex coordinates of the calibrated scale; determining a positioning point of the to-be-detected mark according to the positioning data of the to-be-detected mark; and determining the distance from the positioning point of the to-be-measured mark to the zero graduation line, determining a measurement reading according to the distance, the graduated scale design scheme, the positioning data and the inclination angle of the graduated scale, and determining an installation inspection result according to the measurement reading. By applying the method, automatic inspection of the blade mounting quality can be realized, and the inspection efficiency can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of quality inspection, and in particular to a method for inspecting blade installation quality and related equipment. Background Art

[0002] In the assembly scenario of a wind turbine (i.e., a wind turbine), the wind turbine blades need to be installed on the wind rotor. During the installation process, it is necessary to identify whether the zero scale of the scale representing the blade position is aligned with the zero mark of the bearing to measure whether the blade is installed in place, so as to ensure that the pitch angle set by the wind turbine control program can be achieved by the pitch bearing.

[0003] At present, during the installation process of the blades, the on-site installers determine whether the blades are installed in place, and then take photos of the scene where the blade scale and the bearing zero point mark are located, and archive the photos. Subsequently, the auditors will conduct random inspections of the archived photos to verify the quality of the blade installation.

[0004] In actual application scenarios, the photography skills of on-site personnel are usually uneven, and the quality of the photos taken may be poor. Based on the existing quality inspection methods, auditors are required to visually determine the positional relationship between the blade scale and the bearing zero point mark. When the photo quality is poor, it is difficult for auditors to intuitively determine whether the scales are aligned, and manual inspection through other methods may be required, which takes a lot of time and makes the quality inspection efficiency low. Summary of the invention

[0005] In view of this, an embodiment of the present invention provides a method for inspecting the installation quality of blades to solve the problem that based on the existing quality inspection method, if the quality of photos taken on site is poor, it is difficult for the reviewer to obtain the judgment results intuitively, and a lot of time is required to complete the quality inspection, resulting in low quality inspection efficiency.

[0006] The embodiment of the present invention also provides a device for inspecting the installation quality of blades, so as to ensure the practical realization and application of the above method.

[0007] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0008] A method for inspecting blade installation quality, comprising:

[0009] When the installation quality of the target blade needs to be inspected, an installation scene picture corresponding to the target blade is obtained;

[0010] Input the installation scene picture into a preset image recognition model, and use the image recognition model to perform target recognition on the scale and the mark to be measured in the installation scene picture, and obtain target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm;

[0011] Determining whether the target identification information contains complete target object positioning data;

[0012] If the target identification information contains complete target object positioning data, then obtaining the positioning data corresponding to the scale and the positioning data corresponding to the mark to be measured from the target identification information;

[0013] Determining the inclination angle of the scale and the coordinates of each vertex of the scale according to the positioning data corresponding to the scale;

[0014] Determining a zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale;

[0015] Determine the positioning point of the mark to be measured according to the positioning data corresponding to the mark to be measured;

[0016] Based on the zero scale line and the positioning point of the mark to be measured, a target distance is determined; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line;

[0017] According to the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme, the measurement reading corresponding to the mark to be measured is determined, and the installation inspection result of the target blade is determined according to the measurement reading.

[0018] In the above method, optionally, the positioning data corresponding to the scale includes circumscribed rectangle data of the scale, and determining the inclination angle of the scale according to the positioning data corresponding to the scale includes:

[0019] Determine a first aspect ratio based on the circumscribed rectangle data of the scale; the first aspect ratio is the aspect ratio of the circumscribed rectangle of the scale;

[0020] Based on the scale design, determining a second aspect ratio; the second aspect ratio is the aspect ratio of the scale;

[0021] The inclination angle is calculated according to the first aspect ratio, the second aspect ratio and a preset inclination angle calculation formula, and the calculation result is used as the inclination angle of the scale.

[0022] In the above method, optionally, the process of determining the coordinates of each vertex of the scale includes:

[0023] Based on the circumscribed rectangle data of the scale, determining the coordinates of each vertex of the circumscribed rectangle;

[0024] For each target vertex, among the vertices of the circumscribed rectangle, the vertex located on the same circumscribed rectangle side as the target vertex is used as the associated vertex of the target vertex; the target vertex is the vertex of the scale;

[0025] For each of the target vertices, the length of the line segment between the target vertex and its associated vertices is calculated based on the inclination of the scale and the coordinates of the associated vertices of the target vertex, and the coordinates of the target vertex are calculated based on the lengths of the line segments and the coordinates of the associated vertices of the target vertex to obtain the coordinates of the vertices of the scale.

[0026] In the above method, optionally, determining the target distance based on the zero scale line and the positioning point of the mark to be measured includes:

[0027] Determine a reading line corresponding to the mark to be measured; the reading line is a straight line passing through the positioning point of the mark to be measured and perpendicular to the scale line of the scale;

[0028] Determine the first vertical intercept corresponding to the zero scale line;

[0029] Determine the second vertical intercept corresponding to the reading line;

[0030] The distance between the zero scale line and the reading line is calculated according to the first vertical intercept, the second vertical intercept and the inclination angle of the scale, and the distance is used as the target distance.

[0031] Optionally, the method described above includes determining the measurement reading corresponding to the mark to be measured based on the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme, including:

[0032] Determining a unit scale length corresponding to the scale based on the positioning data corresponding to the scale, the inclination angle of the scale and the scale design scheme;

[0033] The target distance is divided by the unit scale length, and the quotient of the target distance and the unit scale length is taken as the measurement reading.

[0034] In the above method, optionally, determining the installation inspection result of the target blade according to the measurement reading includes:

[0035] Determining whether the measurement reading is within a preset standard reading range;

[0036] If the measured reading is within the standard reading range, a conclusion indicating that the target blade has been correctly installed is taken as an installation inspection result of the target blade;

[0037] If the measured reading is not within the standard reading range, a conclusion indicating that the target blade is not correctly installed is taken as the installation inspection result of the target blade.

[0038] The above method may optionally further include:

[0039] In a case where the target identification information includes complete target object positioning data, copying the installation scene picture to obtain a copy picture corresponding to the installation scene picture;

[0040] Drawing a zero scale auxiliary line corresponding to the zero scale line on the copied image, and drawing a reading auxiliary line on the copied image, to obtain a marked copied image; the reading auxiliary line is a straight line passing through the positioning point of the mark to be measured and perpendicular to the scale line of the scale ruler;

[0041] The marked duplicated image and the installation inspection result of the target blade are used as quality inspection feedback information corresponding to the target blade, and the quality inspection feedback information is sent to the user.

[0042] A blade installation quality inspection device, comprising:

[0043] A picture acquisition unit, used for acquiring an installation scene picture corresponding to the target blade when the installation quality of the target blade needs to be inspected;

[0044] A target recognition unit, used to input the installation scene picture into a preset image recognition model, and the image recognition model performs target recognition on the scale and the mark to be measured in the installation scene picture, and obtains target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm;

[0045] A data determination unit, used to determine whether the target identification information contains complete target object positioning data;

[0046] A data acquisition unit, configured to acquire the positioning data corresponding to the scale and the positioning data corresponding to the mark to be measured from the target identification information if the target identification information contains complete target object positioning data;

[0047] A first determining unit, configured to determine the inclination angle of the scale and the coordinates of each vertex of the scale according to the positioning data corresponding to the scale;

[0048] A second determining unit, configured to determine a zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale;

[0049] A third determining unit, configured to determine a positioning point of the mark to be detected according to the positioning data corresponding to the mark to be detected;

[0050] A fourth determination unit is used to determine a target distance based on the zero scale line and the positioning point of the mark to be measured; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line;

[0051] The fifth determination unit is used to determine the measurement reading corresponding to the mark to be measured based on the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme, and determine the installation inspection result of the target blade based on the measurement reading.

[0052] A storage medium comprises stored instructions, wherein when the instructions are executed, a device where the storage medium is located is controlled to execute the above-mentioned method for inspecting blade installation quality.

[0053] An electronic device comprises a memory and one or more instructions, wherein the one or more instructions are stored in the memory and are configured to be executed by one or more processors to perform the above-mentioned method for inspecting blade installation quality.

[0054] A method for inspecting blade installation quality based on the above-mentioned embodiment of the present invention includes: when it is necessary to inspect the installation quality of a target blade, obtaining an installation scene picture corresponding to the target blade; inputting the installation scene picture into a preset image recognition model, and using the image recognition model to perform target recognition on a scale and a mark to be tested in the installation scene picture, and obtaining target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm; judging whether the target recognition information contains complete target object positioning data; if the target recognition information contains complete target object positioning data, obtaining the positioning data corresponding to the scale from the target recognition information; The method provided by the embodiment of the present invention can be used to identify the target of the installation scene pictures taken on site through an image recognition model based on deep learning, and obtain the positioning data of the mark to be tested and the positioning data of the mark to be tested corresponding to the positioning data of the scale; determine the zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale; determine the positioning point of the mark to be tested based on the positioning data corresponding to the mark to be tested; determine the target distance based on the zero scale line and the positioning point of the mark to be tested; the target distance is the distance from the positioning point of the mark to be tested to the zero scale line; determine the measurement reading corresponding to the mark to be tested based on the target distance, the positioning data corresponding to the scale, the inclination angle of the scale and the preset scale design scheme, and determine the installation inspection result of the target blade based on the measurement reading. By applying the method provided by the embodiment of the present invention, the image recognition model based on deep learning can be used to perform target recognition on the installation scene pictures taken on site, and the positioning data of the scale and the mark to be tested in the installation scene pictures can be obtained. The image recognition model can have good recognition ability for various pictures and can effectively identify the target. Based on the positioning data of the scale and the mark to be measured, the distance between the mark to be measured and the zero scale line of the scale can be further determined, so as to obtain the reading of the current measurement scene, and thus determine whether the blade is installed in place. Based on this process, the automatic inspection of the blade installation quality can be realized without manual review, and all on-site photos can be covered for inspection, which can effectively improve the efficiency of quality inspection and improve the accuracy of quality inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0056] Figure 1 A method flow chart of a method for inspecting blade installation quality provided by an embodiment of the present invention;

[0057] Figure 2 A schematic structural diagram of a scale and its circumscribed rectangle provided by an embodiment of the present invention;

[0058] Figure 3 A schematic structural diagram of a measured mark, a scale and a circumscribed rectangle thereof provided in an embodiment of the present invention;

[0059] Figure 4 A schematic structural diagram of a blade installation quality inspection device provided by an embodiment of the present invention;

[0060] Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] In this application, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0063] The embodiment of the present invention provides a method for inspecting blade installation quality. The method can be applied to a quality inspection system. The execution subject can be a processor or a server of the quality inspection system. The method flow chart of the method is as follows: Figure 1 As shown, including:

[0064] S101: when the installation quality of a target blade needs to be inspected, obtaining an installation scene picture corresponding to the target blade;

[0065] The method provided by the embodiment of the present invention can be applied to the assembly scene of a wind turbine. After the on-site staff assembles the blades onto the wind rotor, they need to face the plane where the scale representing the blade position and the mark representing the zero point of the variable pitch bearing are located to take pictures, that is, they need to take pictures of the scale and the mark. The on-site staff can upload the taken on-site pictures to the quality inspection system, or send them to the quality inspector to upload to the quality inspection system, so as to trigger the quality inspection process of the blade installation.

[0066] In the method provided by the embodiment of the present invention, when the quality inspection system receives a picture uploaded by a user, that is, a picture taken on-site facing the scale and the mark, it can be considered that the installation quality of the blade associated with the picture needs to be inspected, and the blade is used as the target blade. The picture uploaded by the user is used as the installation scene picture corresponding to the target blade.

[0067] S102: Inputting the installation scene picture into a preset image recognition model, allowing the image recognition model to perform target recognition on the scale and the mark to be measured in the installation scene picture, and obtaining target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm;

[0068] In the method provided by the embodiment of the present invention, an image recognition model can be constructed in advance based on a deep learning algorithm. Specifically, sample data can be collected and labeled according to the task of target recognition, so as to obtain the final image recognition model through model training. The task of image recognition is to identify the scale and the mark to be measured. The output parameters of the model can be determined according to actual needs. The circumscribed rectangle of the scale, the deviation of the scale, the circumscribed rectangle of the mark to be measured, etc. can be used as output data, and the inclination angle of the scale, the coordinates of each vertex of the scale, the positioning point of the mark to be measured, etc. can also be further used as output data. The image recognition model can adopt an existing deep learning framework, which can be selected according to actual needs without affecting the function of the method provided by the embodiment of the present invention.

[0069] The quality inspection system can input the current installation scene picture into a pre-built image recognition model, identify the scale and the mark to be tested in the installation scene picture through the image recognition model, and obtain the target recognition information output after processing by the image recognition model.

[0070] S103: Determine whether the target identification information contains complete target object positioning data;

[0071] In the method provided by the embodiment of the present invention, it is possible to determine whether the target recognition information contains complete target object positioning data according to a pre-set image recognition task, that is, to determine whether the image recognition model has recognized the complete positioning information of the scale and the mark to be measured. For example, according to the setting, the image recognition model needs to output the deflection direction (orientation) of the scale, the circumscribed rectangle data of the scale, the circumscribed rectangle data of the mark to be measured, and other contents. If the target recognition information contains these types of data, it is considered that the target recognition information contains complete target object positioning data. If a certain type of data is missing in the target recognition information, it is considered that the target recognition information does not contain complete target object positioning data.

[0072] S104: if the target identification information contains complete target object positioning data, obtaining positioning data corresponding to the scale and positioning data corresponding to the mark to be measured from the target identification information;

[0073] In the method provided by the embodiment of the present invention, if the target identification information contains complete target object positioning data, the positioning data corresponding to the scale and the positioning data corresponding to the mark to be measured are obtained from the target identification information. The positioning data corresponding to the scale can be the deflection direction of the scale, the circumscribed rectangle data of the scale (such as the coordinates of the vertices of the circumscribed rectangle), or the coordinates and inclination of each vertex of the scale, etc. The positioning data corresponding to the mark to be measured can be the circumscribed rectangle data of the measured mark, or the positioning point of the measured mark. It should be noted that the specific parameters in the positioning data of the scale and the mark to be measured are determined by the image recognition task of the model, and the image recognition task of the model can be set according to actual needs.

[0074] In the method provided by the embodiment of the present invention, if the target recognition information does not include complete target object positioning data, response information such as recognition failure can be fed back to the user without performing subsequent processes.

[0075] S105: determining the inclination angle of the scale and the coordinates of each vertex of the scale according to the positioning data corresponding to the scale;

[0076] In the method provided by the embodiment of the present invention, the inclination angle of the ruler in the installation scene picture, that is, the inclination angle of the ruler, and the coordinates of each vertex of the ruler can be determined based on the positioning data of the ruler. For example, the inclination angle of the ruler and the coordinates of each vertex can be identified by an image recognition model, that is, the positioning data corresponding to the ruler can contain these data, and can be directly obtained from the positioning data. For another example, the correlation between the posture of the ruler and the inclination angle and the vertex coordinates can be analyzed in advance, and the calculation method of the inclination angle and the coordinates of each vertex can be set. In the quality inspection process, the inclination angle of the ruler and the coordinates of each vertex can be calculated according to the positioning data of the ruler through the corresponding calculation method.

[0077] S106: Determine a zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale;

[0078] In the method provided in the embodiment of the present invention, the zero scale point of the scale in the installation scene picture can be located based on the inclination angle of the scale and the coordinates of each vertex, and a straight line perpendicular to the scale line (that is, the long side of the scale) is drawn through the zero scale point to obtain the zero scale line. It can be understood that the design of the scale is known, so the coordinates of the zero scale point can be located according to the current posture of the scale.

[0079] S107: Determine the positioning point of the mark to be measured according to the positioning data corresponding to the mark to be measured;

[0080] In the method provided by the embodiment of the present invention, the positioning point of the mark to be measured can be determined based on the positioning data of the mark to be measured. For example, if the positioning data corresponding to the mark to be measured is a circumscribed rectangle of the mark to be measured, the center point of the circumscribed rectangle can be used as the positioning point of the mark to be measured.

[0081] S108: Determine a target distance based on the zero scale line and the positioning point of the mark to be measured; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line;

[0082] In the method provided in the embodiment of the present invention, the distance from the positioning point of the mark to be measured to the zero scale line can be calculated according to a preset distance calculation method, and the distance is used as the target distance. It can be understood that the positioning point and the zero scale line of the mark to be measured are currently known, and determining the target distance is to determine the distance from a certain point to a certain straight line. Various methods of solving the distance from the point to the straight line can be used to achieve the solution, and the specific calculation method used can be configured according to actual needs.

[0083] S109: Determine the measurement reading corresponding to the mark to be measured according to the target distance, the positioning data corresponding to the scale, the inclination angle of the scale and a preset scale design scheme, and determine the installation inspection result of the target blade according to the measurement reading.

[0084] In the method provided by the embodiment of the present invention, the unit scale of the scale can be obtained from the design scheme of the scale, thereby converting the reading of the target distance, that is, determining the reading represented by the target distance in the current measurement scenario, and using the conversion result as the measurement reading corresponding to the mark to be measured. According to the measurement reading, it is determined whether the target blade is currently installed in place, thereby obtaining the installation inspection result. The installation inspection result can be fed back to the user through the front end.

[0085] Based on the method provided by the embodiment of the present invention, when it is necessary to inspect the installation quality of the target blade, obtain the installation scene picture corresponding to the target blade; input the installation scene picture into a preset image recognition model, and the image recognition model performs target recognition on the scale and the mark to be tested in the installation scene picture, and obtains the target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm; it is determined whether the target recognition information contains complete target object positioning data; if the target recognition information contains complete target object positioning data, the positioning data corresponding to the scale and the mark to be tested are obtained from the target recognition information Corresponding positioning data; according to the positioning data corresponding to the scale, determine the inclination of the scale and the coordinates of each vertex of the scale; based on the inclination of the scale and the coordinates of each vertex, determine the zero scale line corresponding to the scale; according to the positioning data corresponding to the mark to be measured, determine the positioning point of the mark to be measured; based on the zero scale line and the positioning point of the mark to be measured, determine the target distance; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line; according to the target distance, the positioning data corresponding to the scale, the inclination of the scale and the preset scale design scheme, determine the measurement reading corresponding to the mark to be measured, and determine the installation inspection result of the target blade based on the measurement reading. Applying the method provided in the embodiment of the present invention, the image recognition model based on deep learning can be used to perform target recognition on the installation scene pictures taken on site, and the positioning data of the scale and the mark to be measured in the installation scene pictures can be obtained. The image recognition model can have good recognition ability for various types of pictures and can effectively identify the target. Based on the positioning data of the scale and the mark to be measured, the distance between the mark to be measured and the zero scale line of the scale can be further determined, so as to obtain the reading of the current measurement scene, and thus determine whether the blade is installed in place. Based on this process, the automatic inspection of the blade installation quality can be realized without manual review, and all on-site photos can be covered for inspection, which can effectively improve the efficiency of quality inspection and improve the accuracy of quality inspection.

[0086] exist Figure 1 On the basis of the method shown, in the method provided by the embodiment of the present invention, the positioning data corresponding to the scale mentioned in step S104 includes the circumscribed rectangle data of the scale, and the process of determining the inclination angle of the scale according to the positioning data corresponding to the scale mentioned in step S105 includes:

[0087] Determine a first aspect ratio based on the circumscribed rectangle data of the scale; the first aspect ratio is the aspect ratio of the circumscribed rectangle of the scale;

[0088] In the method provided in the embodiment of the present invention, the positioning data of the ruler includes the circumscribed rectangle data of the ruler, which may include the coordinates of each vertex of the circumscribed rectangle, and may also include the aspect ratio of the circumscribed rectangle. If the data output by the image recognition model includes the aspect ratio of the circumscribed rectangle of the ruler, then the aspect ratio output by the model can be used as the first aspect ratio. If the circumscribed rectangle data of the ruler does not include the aspect ratio of the circumscribed rectangle, then based on the coordinates of each vertex of the circumscribed rectangle, the length and width of the circumscribed rectangle can be calculated, and then the aspect ratio of the circumscribed rectangle can be calculated, and the calculated aspect ratio can be used as the first aspect ratio.

[0089] Based on the scale design, determining a second aspect ratio; the second aspect ratio is the aspect ratio of the scale;

[0090] In the method provided in the embodiment of the present invention, the design length and design width of the ruler can be obtained from the design scheme of the ruler, and the aspect ratio of the ruler can be calculated according to the design length and design width, and the aspect ratio of the ruler can be used as the second aspect ratio.

[0091] The inclination angle is calculated according to the first aspect ratio, the second aspect ratio and a preset inclination angle calculation formula, and the calculation result is used as the inclination angle of the scale.

[0092] In the method provided in the embodiment of the present invention, a calculation formula for the inclination angle of the scale can be set in advance through structural analysis, that is, the so-called inclination angle calculation formula. The inclination angle calculation formula in the embodiment of the present invention uses the length-to-width ratio of the scale and the length-to-width ratio of the circumscribed rectangle of the scale to realize the inclination angle calculation. Therefore, the first length-to-width ratio and the second length-to-width ratio can be substituted into the inclination angle calculation formula for calculation to obtain a calculation result, that is, the inclination angle of the scale. The inclination angle calculation formula in the embodiment of the present invention can be as follows:

[0093] θ = acrtan[(m - n) / (1 - m·n)] (Formula 1)

[0094] Wherein, θ represents the inclination angle of the scale, m represents the aspect ratio of the circumscribed rectangle of the scale (ie, the first aspect ratio), and n represents the aspect ratio of the scale (ie, the second aspect ratio).

[0095] Next, we briefly explain the derivation process of Formula 1. Figure 2 A structural schematic diagram of a scale and its circumscribed rectangle is shown. Figure 2In the equation, the left-leaning rectangle is the ruler, n represents the aspect ratio of the ruler, x represents the width of the ruler, nx is the product of n and x, and represents the length of the ruler. m represents the aspect ratio of the circumscribed rectangle, y represents the width of the circumscribed rectangle, my is the product of m and y, and represents the length of the circumscribed rectangle. θ2 is the inclination angle of the ruler, which is θ in equation 1. H1 represents the length from the lower right vertex of the ruler to the lower right vertex of the circumscribed rectangle, H2 represents the length from the lower right vertex of the ruler to the upper right vertex of the circumscribed rectangle, W1 represents the length from the lower left vertex of the ruler to the lower right vertex of the circumscribed rectangle, and W2 represents the length from the upper right vertex of the ruler to the upper right vertex of the circumscribed rectangle. It can be seen that:

[0096] sin(θ2)=W2 / nx (Formula 2)

[0097] sin(θ2)=H1 / x (Equation 3)

[0098] cos(θ2)=H2 / nx (Equation 4)

[0099] cos(θ2)=W1 / x (Equation 5)

[0100] H1+H2=my (Formula 6)

[0101] W1+W2=y (Formula 7)

[0102] By changing equation 2, equation 3, equation 4 and equation 5, we can get:

[0103] W2=nx·sin(θ2) (Formula 8)

[0104] H1=x·sin(θ2) (Formula 9)

[0105] H2=nx·cos(θ2) (Formula 10)

[0106] W1=x·cos(θ2) (Formula 11)

[0107] Substituting equations 8 to 11 into equations 6 and 7, we can obtain:

[0108] x·sin(θ2)+nx·cos(θ2)=my (Equation 12)

[0109] x·cos(θ2)+nx·sin(θ2)=y (Equation 13)

[0110] Based on equation 12 and equation 13, we can get:

[0111] m=(sin(θ2)+n·cos(θ2)) / (cos(θ2)+n·sin(θ2)) (Equation 14)

[0112] Since m and n are known, we can solve θ2 (that is, θ in Equation 1) based on Equation 14 to obtain Equation 1.

[0113] It should be noted that Figure 2 This is only an example of a case where a scale is tilted to the left provided by the method provided by an embodiment of the present invention in order to better illustrate the example. Regardless of the postures of the scale and the circumscribed rectangle in actual application scenarios, the tilt angle can be solved based on a similar principle.

[0114] Based on the method provided in the embodiment of the present invention, the image recognition model only needs to output the circumscribed rectangle data of the scale, and the scale inclination is obtained through data calculation based on the circumscribed rectangle data, which is conducive to improving the accuracy of inclination calculation.

[0115] Based on the method provided in the above embodiment, the process of determining the coordinates of each vertex of the scale mentioned in step S104 includes:

[0116] Based on the circumscribed rectangle data of the scale, determining the coordinates of each vertex of the circumscribed rectangle;

[0117] In the method provided by the embodiment of the present invention, the coordinates of each vertex of the circumscribed rectangle of the scale can be obtained from the circumscribed rectangle data of the scale.

[0118] For each target vertex, among the vertices of the circumscribed rectangle, the vertex located on the same circumscribed rectangle side as the target vertex is used as the associated vertex of the target vertex; the target vertex is the vertex of the scale;

[0119] In the method provided by the embodiment of the present invention, for each vertex of the scale, the vertex of the circumscribed rectangle on the edge of the circumscribed rectangle where the vertex is located is used as the associated vertex of the vertex.

[0120] For each of the target vertices, the length of the line segment between the target vertex and its associated vertices is calculated based on the inclination of the scale and the coordinates of the associated vertices of the target vertex, and the coordinates of the target vertex are calculated based on the lengths of the line segments and the coordinates of the associated vertices of the target vertex to obtain the coordinates of the vertices of the scale.

[0121] In the method provided in the embodiment of the present invention, for each vertex of the scale, the length of the line segment between the vertex and its associated vertex can be calculated according to the inclination angle of the scale and the coordinates of the vertex associated with the vertex. Figure 2As shown, H1, H2, W1 and W2 are the lengths of the line segments between the vertices of the scale and the corresponding vertices of the circumscribed rectangle, which can be calculated using the above formulas 8 to 11. After obtaining the lengths of the line segments between each vertex of the scale and its associated vertex, the coordinates of each vertex of the scale can be calculated based on the coordinates of each vertex of the circumscribed rectangle. For example, Figure 2 In the structure shown, regarding the upper right corner vertex of the ruler, the line segment length W2 between the vertex and the upper right corner vertex of the circumscribed rectangle is calculated. The coordinates of the upper right corner vertex of the circumscribed rectangle are known to be (a, b), so the coordinates of the upper right corner vertex of the ruler are (a-W2, b).

[0122] Based on the method provided in the embodiment of the present invention, an image recognition model is used in which only the circumscribed rectangle data of the scale is output, and based on the circumscribed rectangle data and the inclination angle, the coordinates of each vertex of the scale are obtained through data calculation, which is conducive to improving the accuracy of the coordinates of the vertex of the scale.

[0123] exist Figure 1 On the basis of the method shown, in the method provided by the embodiment of the present invention, the process of determining the target distance based on the zero scale line and the positioning point of the mark to be measured mentioned in step S108 includes:

[0124] Determine a reading line corresponding to the mark to be measured; the reading line is a straight line passing through the positioning point of the mark to be measured and perpendicular to the scale line of the scale;

[0125] In the method provided by the embodiment of the present invention, a straight line perpendicular to the scale line (that is, the long side of the scale) can be drawn through the positioning point of the mark to be measured, and the straight line can be used as the reading line of the mark to be measured.

[0126] Determine the first vertical intercept corresponding to the zero scale line;

[0127] Determine the second vertical intercept corresponding to the reading line;

[0128] In the method provided by the embodiment of the present invention, the vertical intercept of the zero scale line can be determined according to the equation of the zero scale line, and the vertical intercept can be used as the first vertical intercept. According to the equation of the reading line, the vertical intercept of the reading line can be determined, and the vertical intercept can be used as the second vertical intercept. It can be understood that the zero scale line is a straight line passing through the zero scale point and perpendicular to the scale lines of the scale. On the basis of determining the zero scale point, combined with the inclination of the scale, the equation of the zero scale line can be obtained. Similarly, the equation of the reading line can be obtained.

[0129] The distance between the zero scale line and the reading line is calculated according to the first vertical intercept, the second vertical intercept and the inclination angle of the scale, and the distance is used as the target distance.

[0130] In the method provided in the embodiment of the present invention, based on the first vertical intercept and the second vertical intercept, the intercept difference between the zero scale line and the reading line can be calculated, and combined with the intercept difference, the distance between the zero scale line and the reading line can be calculated, and the distance is the target distance. In the embodiment of the present invention, the calculation principle of the target distance can be as follows:

[0131] S = cos(θ)·|z1-z2| (Equation 15)

[0132] Among them, S represents the target distance, θ is the inclination angle of the scale, z1 represents the first vertical intercept (i.e. the vertical intercept of the zero scale line), and z2 represents the second vertical intercept (i.e. the vertical intercept of the reading line).

[0133] In order to better illustrate the method provided by the embodiment of the present invention, the derivation process of the principle of calculating the target distance is briefly described below. Figure 2 Based on Figure 3 The structure shown is used as an example. Figure 3 The θ in is equivalent to Figure 2 θ2 in Figure 3 and Figure 2 The same symbols have the same meaning. Figure 3 In the figure, the mark similar to "S" in the lower left rectangle is the mark to be measured, k (i.e. K) represents a scale of the ruler, the straight line L1-L is the zero scale line, and the straight line L1'-L' is the reading line. It can be understood that the target distance is the length of the line segment C-C', that is, the length of the line segment P-L2'. The length of the line segment P-L2' is equal to the product of cos(θ) and the length of the line segment L2-L2', and the line segment L2-L2' can be understood as a line segment formed by the intersection of a vertical line with the zero scale line and the reading line, and the length of the line segment is the difference between the vertical intercept of the zero scale line and the vertical intercept of the reading line. It should be noted that Figure 3 The scale on the ruler is shown, and the scale illustration is only for explaining the meaning of k. In actual application scenarios, whether the image contains a scale illustration does not affect the function of the method provided in the embodiment of the present invention.

[0134] exist Figure 1 On the basis of the method shown in the figure, in the method provided in the embodiment of the present invention, the process of determining the measurement reading corresponding to the mark to be measured according to the target distance, the positioning data corresponding to the scale, the inclination angle of the scale and the preset scale design scheme mentioned in step S109 includes:

[0135] Determining a unit scale length corresponding to the scale based on the positioning data corresponding to the scale, the inclination angle of the scale and the scale design scheme;

[0136] In the method provided by the embodiment of the present invention, the positioning data corresponding to the ruler includes the circumscribed rectangle data of the ruler, and the length and width of the circumscribed rectangle of the ruler can be obtained. The ruler design scheme records the design information of the ruler, including the ratio of the unit scale of the ruler to the total length of the ruler, and the ratio is expressed as kr. Based on the above formulas 1 to 15, it can be known that:

[0137] kr=k / (nx) (Equation 16)

[0138] k=kr·nx (Equation 17)

[0139] In the method provided in the embodiment of the present invention, the unit scale length corresponding to the scale ruler can be calculated according to the following formula.

[0140] k=(kr·n·my) / (sin(θ)+n·cos(θ)) (Equation 18)

[0141] Among them, k is the unit scale length, and the meanings of the other parameters are the same as the corresponding parameters in the previous formula.

[0142] In order to better illustrate the method provided by the embodiment of the present invention, the principle of calculating the unit scale length is briefly described below. Figure 2 , Figure 3 , based on the above formulas 1 to 15, based on formula 12, we can get:

[0143] x=my / (sin(θ)+n·cos(θ)) (Equation 19)

[0144] nx=(n·my) / (sin(θ)+n·cos(θ)) (Equation 20)

[0145] Substituting equation 20 into equation 17 and integrating them, we can get equation 18.

[0146] The target distance is divided by the unit scale length, and the quotient of the target distance and the unit scale length is taken as the measurement reading.

[0147] In the method provided in the embodiment of the present invention, the quotient obtained by dividing the target distance by the unit scale length is used as the measurement reading. Figure 3 Based on the structure shown, the measurement readings can be calculated as follows:

[0148] Measurement reading = [cos(θ)·L2L2'] / [(kr·n·my) / (sin(θ)+n·cos(θ))].

[0149] It can be understood that the unit of the measurement reading corresponds to the unit scale of the ruler. For example, if the unit scale of the ruler represents 1 cm and the value of the measurement reading is 1.5, then it represents 1.5 cm.

[0150] exist Figure 1 On the basis of the method shown, in the method provided by the embodiment of the present invention, the process of determining the installation inspection result of the target blade according to the measurement reading mentioned in step S109 includes:

[0151] Determining whether the measurement reading is within a preset standard reading range;

[0152] In the method provided in the embodiment of the present invention, a standard reading range can be set in advance according to actual needs, that is, a reading range that characterizes the installation of the blade. The current measurement reading can be compared with the upper and lower limits of the standard reading range to determine whether the current measurement reading is within the preset standard reading range.

[0153] If the measured reading is within the standard reading range, a conclusion indicating that the target blade has been correctly installed is taken as an installation inspection result of the target blade;

[0154] In the method provided by the embodiment of the present invention, if the current measurement reading is within the standard reading range, it is determined that the installation inspection result of the target blade is that the blade has been correctly installed.

[0155] If the measured reading is not within the standard reading range, a conclusion indicating that the target blade is not correctly installed is taken as the installation inspection result of the target blade.

[0156] In the method provided by the embodiment of the present invention, if the current measurement reading exceeds the standard reading range, it is determined that the installation inspection result of the target blade is that the blade is not correctly installed.

[0157] Based on the method provided in the embodiment of the present invention, whether the measured reading indicates that the blade has been correctly installed can be measured by setting a standard reading range, thereby avoiding misjudgment caused by errors and facilitating further improvement of the accuracy of the inspection.

[0158] exist Figure 1 On the basis of the method shown, the method provided in the embodiment of the present invention further includes:

[0159] In a case where the target identification information includes complete target object positioning data, copying the installation scene picture to obtain a copy picture corresponding to the installation scene picture;

[0160] In the method provided by the embodiment of the present invention, when the target identification information includes complete target object positioning data, the installation scene picture can be copied so as to use the copied picture for drawing operation.

[0161] Drawing a zero scale auxiliary line corresponding to the zero scale line on the copied image, and drawing a reading auxiliary line on the copied image, to obtain a marked copied image; the reading auxiliary line is a straight line passing through the positioning point of the mark to be measured and perpendicular to the scale line of the scale ruler;

[0162] In the method provided by the embodiment of the present invention, the position of the scale and the position of the mark to be measured can be located on the copied image according to the positioning data of the scale and the positioning data of the mark to be measured, and a zero scale auxiliary line corresponding to the zero scale line can be drawn at the position of the scale, that is, a zero scale line can be drawn on the scale object on the copied image, and a straight line passing through the positioning point of the mark to be measured and perpendicular to the scale line of the scale, that is, a reading auxiliary line, can be drawn according to the position of the mark to be measured on the copied image. Figure 3 As shown, that is, on the copied picture, draw auxiliary lines similar to straight lines L1-L, L1'-L', etc.

[0163] The marked duplicated image and the installation inspection result of the target blade are used as quality inspection feedback information corresponding to the target blade, and the quality inspection feedback information is sent to the user.

[0164] In the method provided by an embodiment of the present invention, the copied image with auxiliary lines drawn and the installation inspection result of the target blade are used together as quality inspection feedback information of the target blade. The quality inspection feedback information is sent to the user through the front end so that the user can obtain the quality inspection result and perform secondary verification based on the marked copied image.

[0165] Based on the method provided in the embodiment of the present invention, corresponding auxiliary lines can be drawn in the installation scene picture and fed back to the user together with the installation inspection results, so that the user can perform secondary verification, which is conducive to further improving the accuracy of quality inspection.

[0166] In order to better illustrate the method provided by the embodiment of the present invention, based on the methods provided by the above embodiments and combined with actual application scenarios, the embodiment of the present invention provides another method for inspecting the installation quality of blades. The method provided by the embodiment of the present invention can be deployed as a mobile application. During the development process, for the task of image recognition, sample data is collected, and the scales and marks in the sample data are calibrated. Based on the labeled data, an image recognition model is constructed using deep learning technology.

[0167] Users (on-site installers) can use the application on their mobile devices to take photos of the site directly after completing the blade assembly operation and upload them to the server. In the case of a network environment, the results returned by the server can be quickly obtained. In the absence of a network environment, the edge computing capabilities of the mobile device can be used to obtain inspection results based on edge computing.

[0168] After the server or edge computing processor receives the installation scene picture uploaded by the user, the image recognition model can perform target detection and output normalized results.

[0169] If the model recognizes that there is a ruler in the image, the model will return the coordinates of the ruler's circumscribed rectangle and the ruler's orientation (the ruler's orientation is divided into left bias, right bias, and no bias).

[0170] If the model recognizes that there is a mark to be tested in the image, the model will return the bounding rectangle of the mark to be tested.

[0171] If the model fails to recognize the scale or the mark to be measured, a response message indicating the recognition failure is fed back to the user.

[0172] When the model recognizes the relevant information of the scale and the mark to be measured, the specific inclination angle and the coordinates of each vertex of the scale are calculated based on the positioning data of the scale using the calculation principles shown in Formulas 1 to 14 above.

[0173] Then, based on the circumscribed rectangle data of the scale, the positioning data of the mark to be measured, and the inclination angle and vertex coordinates of the scale, the measurement reading corresponding to the mark to be measured can be calculated based on the calculation principles shown in equations 15 to 20 above.

[0174] If the above processing is handled by a server, the server returns the measurement readings to the mobile device.

[0175] The mobile device compares the measured reading with the standard reading range according to the standard reading range pre-stored in the application to obtain the inspection result of the blade installation quality. The result is packaged into a human-understandable form and provided to the user through preset rules. Furthermore, it can also draw a similar Figure 3 The auxiliary lines shown will feed back the marked copy version image to the user to facilitate the user's secondary verification.

[0176] Furthermore, in a specific implementation process, the image recognition model can be used to identify the inclination of the output scale through model training, and subsequent calculations can be performed accordingly. The image recognition model can also be used to identify the inclination of the output scale and the coordinates of each vertex of the scale through model training, and subsequent calculations can be performed accordingly. By using the image recognition model to identify the inclination of the scale and the coordinates of the vertices, data calculations can be reduced and the complexity of the program can be reduced.

[0177] and Figure 1 Corresponding to the method for inspecting the installation quality of a blade shown in FIG. 1 , an embodiment of the present invention further provides an apparatus for inspecting the installation quality of a blade. Figure 1 The specific implementation of the method shown in is shown in the structural diagram Figure 4 As shown, including:

[0178] The picture acquisition unit 201 is used to acquire the installation scene picture corresponding to the target blade when the installation quality of the target blade needs to be inspected;

[0179] The target recognition unit 202 is used to input the installation scene picture into a preset image recognition model, and the image recognition model performs target recognition on the scale and the mark to be measured in the installation scene picture, and obtains the target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm;

[0180] A data determination unit 203 is used to determine whether the target identification information contains complete target object positioning data;

[0181] A data acquisition unit 204 is configured to acquire the positioning data corresponding to the scale and the positioning data corresponding to the mark to be measured from the target identification information if the target identification information contains complete target object positioning data;

[0182] A first determining unit 205, configured to determine the inclination angle of the scale and the coordinates of each vertex of the scale according to the positioning data corresponding to the scale;

[0183] A second determining unit 206, configured to determine a zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale;

[0184] A third determining unit 207 is used to determine the positioning point of the mark to be detected according to the positioning data corresponding to the mark to be detected;

[0185] The fourth determining unit 208 is used to determine a target distance based on the zero scale line and the positioning point of the mark to be measured; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line;

[0186] The fifth determination unit 209 is used to determine the measurement reading corresponding to the mark to be measured according to the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme, and determine the installation inspection result of the target blade according to the measurement reading.

[0187] By using the device provided by the embodiment of the present invention, it is possible to perform target recognition on the installation scene pictures taken on site through an image recognition model based on deep learning, and obtain the positioning data of the scale and the mark to be measured in the installation scene pictures. The image recognition model can have good recognition capabilities for various types of pictures and can effectively identify the target. Based on the positioning data of the scale and the mark to be measured, the distance between the mark to be measured and the zero scale line of the scale can be further determined, thereby obtaining the reading of the current measurement scene, and based on this, it can be determined whether the blade is installed in place. Based on this process, the automatic inspection of the blade installation quality can be realized without manual review, and all on-site photos can be covered for inspection, which can effectively improve the efficiency of quality inspection and improve the accuracy of quality inspection.

[0188] exist Figure 4 Based on the device shown, the device provided in the embodiment of the present invention can be further expanded into multiple units. The function of each unit can be found in the description of each embodiment provided in the previous text for the inspection method of blade installation quality, and no further examples will be given here.

[0189] An embodiment of the present invention further provides a storage medium, which includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the above-mentioned method for inspecting the blade installation quality.

[0190] The embodiment of the present invention further provides an electronic device, the structural diagram of which is shown in FIG. Figure 5 As shown, it specifically includes a memory 301 and one or more instructions 302, wherein the one or more instructions 302 are stored in the memory 301 and are configured to be executed by one or more processors 303 to perform the following operations:

[0191] When the installation quality of the target blade needs to be inspected, an installation scene picture corresponding to the target blade is obtained;

[0192] Input the installation scene picture into a preset image recognition model, and use the image recognition model to perform target recognition on the scale and the mark to be measured in the installation scene picture, and obtain target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm;

[0193] Determining whether the target identification information contains complete target object positioning data;

[0194] If the target identification information contains complete target object positioning data, then obtaining the positioning data corresponding to the scale and the positioning data corresponding to the mark to be measured from the target identification information;

[0195] Determining the inclination angle of the scale and the coordinates of each vertex of the scale according to the positioning data corresponding to the scale;

[0196] Determining a zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale;

[0197] Determine the positioning point of the mark to be measured according to the positioning data corresponding to the mark to be measured;

[0198] Based on the zero scale line and the positioning point of the mark to be measured, a target distance is determined; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line;

[0199] According to the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme, the measurement reading corresponding to the mark to be measured is determined, and the installation inspection result of the target blade is determined according to the measurement reading.

[0200] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0201] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0202] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for inspecting blade installation quality, characterized in that: include: When the installation quality of the target blade needs to be inspected, an installation scene picture corresponding to the target blade is obtained; Input the installation scene picture into a preset image recognition model, and use the image recognition model to perform target recognition on the scale and the mark to be measured in the installation scene picture, and obtain target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm; Determining whether the target identification information contains complete target object positioning data; If the target identification information contains complete target object positioning data, then obtaining the positioning data corresponding to the scale and the positioning data corresponding to the mark to be measured from the target identification information; Determining the inclination angle of the scale and the coordinates of each vertex of the scale according to the positioning data corresponding to the scale; Determining a zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale; Determine the positioning point of the mark to be measured according to the positioning data corresponding to the mark to be measured; Based on the zero scale line and the positioning point of the mark to be measured, a target distance is determined; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line; According to the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme, the measurement reading corresponding to the mark to be measured is determined, and the installation inspection result of the target blade is determined according to the measurement reading.

2. The method for inspecting blade installation quality according to claim 1, characterized in that: The positioning data corresponding to the scale includes circumscribed rectangle data of the scale, and determining the inclination angle of the scale according to the positioning data corresponding to the scale includes: Determine a first aspect ratio based on the circumscribed rectangle data of the scale; the first aspect ratio is the aspect ratio of the circumscribed rectangle of the scale; Based on the scale design, determining a second aspect ratio; the second aspect ratio is the aspect ratio of the scale; The inclination angle is calculated according to the first aspect ratio, the second aspect ratio and a preset inclination angle calculation formula, and the calculation result is used as the inclination angle of the scale.

3. The method for inspecting blade installation quality according to claim 2, characterized in that: The process of determining the coordinates of each vertex of the scale comprises: Based on the circumscribed rectangle data of the scale, determining the coordinates of each vertex of the circumscribed rectangle; For each target vertex, among the vertices of the circumscribed rectangle, the vertex located on the same circumscribed rectangle side as the target vertex is used as the associated vertex of the target vertex; the target vertex is the vertex of the scale; For each of the target vertices, the length of the line segment between the target vertex and its associated vertices is calculated based on the inclination of the scale and the coordinates of the associated vertices of the target vertex, and the coordinates of the target vertex are calculated based on the lengths of the line segments and the coordinates of the associated vertices of the target vertex to obtain the coordinates of the vertices of the scale.

4. The method for inspecting blade installation quality according to claim 1, characterized in that: The step of determining the target distance based on the zero scale line and the positioning point of the mark to be measured comprises: Determine a reading line corresponding to the mark to be measured; the reading line is a straight line passing through the positioning point of the mark to be measured and perpendicular to the scale line of the scale; Determine the first vertical intercept corresponding to the zero scale line; Determine the second vertical intercept corresponding to the reading line; The distance between the zero scale line and the reading line is calculated according to the first vertical intercept, the second vertical intercept and the inclination angle of the scale, and the distance is used as the target distance.

5. The method for inspecting blade installation quality according to claim 1, characterized in that: Determining the measurement reading corresponding to the mark to be measured based on the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme includes: Determining a unit scale length corresponding to the scale based on the positioning data corresponding to the scale, the inclination angle of the scale and the scale design scheme; The target distance is divided by the unit scale length, and the quotient of the target distance and the unit scale length is taken as the measurement reading.

6. The method for inspecting blade installation quality according to claim 1, characterized in that: Determining the installation inspection result of the target blade according to the measurement reading includes: Determining whether the measurement reading is within a preset standard reading range; If the measured reading is within the standard reading range, a conclusion indicating that the target blade has been correctly installed is taken as an installation inspection result of the target blade; If the measured reading is not within the standard reading range, a conclusion indicating that the target blade is not correctly installed is taken as the installation inspection result of the target blade.

7. The method for inspecting blade installation quality according to claim 1, characterized in that: Also includes: In a case where the target identification information includes complete target object positioning data, copying the installation scene picture to obtain a copy picture corresponding to the installation scene picture; Drawing a zero scale auxiliary line corresponding to the zero scale line on the copied image, and drawing a reading auxiliary line on the copied image, to obtain a marked copied image; The reading auxiliary line is a straight line passing through the positioning point of the mark to be measured and perpendicular to the scale line of the scale; The marked duplicated image and the installation inspection result of the target blade are used as quality inspection feedback information corresponding to the target blade, and the quality inspection feedback information is sent to the user.

8. A blade installation quality inspection device, characterized in that: include: A picture acquisition unit, used for acquiring an installation scene picture corresponding to the target blade when the installation quality of the target blade needs to be inspected; A target recognition unit, used to input the installation scene picture into a preset image recognition model, and the image recognition model performs target recognition on the scale and the mark to be measured in the installation scene picture, and obtains target recognition information output by the image recognition model; the image recognition model is a model built based on a deep learning algorithm; A data determination unit, used to determine whether the target identification information contains complete target object positioning data; A data acquisition unit, configured to acquire the positioning data corresponding to the scale and the positioning data corresponding to the mark to be measured from the target identification information if the target identification information contains complete target object positioning data; A first determining unit, configured to determine the inclination angle of the scale and the coordinates of each vertex of the scale according to the positioning data corresponding to the scale; A second determining unit, configured to determine a zero scale line corresponding to the scale based on the inclination angle of the scale and the coordinates of each vertex of the scale; A third determining unit, configured to determine a positioning point of the mark to be detected according to the positioning data corresponding to the mark to be detected; A fourth determination unit is used to determine a target distance based on the zero scale line and the positioning point of the mark to be measured; the target distance is the distance from the positioning point of the mark to be measured to the zero scale line; The fifth determination unit is used to determine the measurement reading corresponding to the mark to be measured based on the target distance, the positioning data corresponding to the scale, the inclination of the scale and a preset scale design scheme, and determine the installation inspection result of the target blade based on the measurement reading.

9. A storage medium, characterized in that: The storage medium includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to execute the method for inspecting the blade installation quality according to any one of claims 1 to 7.

10. An electronic device, characterized in that: The device comprises a memory and one or more instructions, wherein the one or more instructions are stored in the memory and configured to be executed by one or more processors to perform the method for inspecting the installation quality of a blade according to any one of claims 1 to 7.

Images