JP2016136111A - Shape error determination device and determination result image generation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a shape error determination device capable of properly determining a various types of shape errors, and a determination result image generation device provided with the shape error determination device and capable of generating a highly visible image as a determination result image that is to be used when a shape error determination result is presented.MEANS FOR SOLVING THE PROBLEM: A shape error determination device 10 comprises: a first-shape abnormality determination part (defective error determination unit 12) that compares a size of a first-shape portion (defect portion) detected by a first-shape detection part (defect detection unit 22) with a first-shape abnormality determination threshold (defective error determination threshold) to determine whether or not the first-shape portion is abnormal; and a second-shape abnormality determination part (abrasion error determination unit 11) that determines whether or not a second-shape portion (abrasion portion) is abnormal on the basis of three-dimensional shape information on a region in a predetermined surface (abrasion surface) in which the first-shape portion (defect portion) is not formed.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a shape error determination device and a determination result image generation device including the shape error determination device.

A method for measuring the three-dimensional shape of an object using a twin-lens camera, a laser scanner, ultrasonic waves, or the like is known. In the case of a twin-lens camera, a captured image is used, and in the case of a laser scanner or the like, a camera is prepared separately, and a captured image of an object is acquired simultaneously with measurement, and the measurement result is superimposed on the captured image and displayed. It is possible to display a result that makes it easy to determine the correspondence between the measurement result and the measurement location on the object. When determining the presence or absence of an object shape error based on the measured three-dimensional shape of the object, the shape error determination result and the object are displayed by superimposing the shape error determination result on the captured image. As a result, it is possible to display a result that makes it easy to determine the correspondence with the portion where the shape error is detected (that is, the portion determined to have a shape error).
For example, in Patent Document 1, a target region in an object is detected from a captured image acquired by a monocular camera, and the detected result is displayed in a superimposed manner on the captured image, so that the detected target region is easy to understand and the target region Has disclosed a technique for displaying a result that can be easily discriminated when there is an error in detection.

In 3D shape measurement, there are cases where it is desired to determine the presence or absence of multiple types of shape errors such as wear errors and missing errors, and there are cases where there are areas where the 3D shape cannot be measured depending on the measurement method and the condition of the object. is there. When displaying results, it is possible to display judgment results for multiple types of shape errors, and to display a warning by detecting a region where a three-dimensional shape cannot be measured and displaying a warning. it can.
One application example of the three-dimensional shape measurement is a wear amount inspection of a contact portion with an overhead line in a pantograph for electric power supply of a train. For example, in Patent Document 2, a large deficiency (deterioration) on the side of a pantograph, which is an area where erroneous measurement is likely to occur in a normal pantograph wear amount measurement method, is detected using a luminance difference, and the damage is measured when the wear amount is measured. A technique for reducing erroneous measurement of the amount of wear by measuring the portion is disclosed.

JP 2013-124983 A JP 2006-118900 A

In the pantograph, not only wear but also a defect may occur, and it is necessary to detect the defect part. However, in the above-mentioned Patent Document 2, a defect portion (deterioration portion) is detected as a region where the wear amount of the pantograph cannot be measured, and the measurement is performed by removing the defect portion when measuring the wear amount. Cannot detect or measure the size of the missing part. Therefore, if you want to know not only the presence / absence of wear and the amount of wear, but also the presence / absence of the defect and the size of the defect, you can actually see the object or view the image of the object to check for the presence / absence of the defect. It is necessary to determine the size of the image, but this determination is difficult.
Moreover, in the said patent document 2, in order to detect a defect | deletion part with a luminance difference, it is necessary to set so that the light quantity which illuminates a pantograph may become the same by each measurement. In other words, it is not possible to cope with outdoor measurement that is affected by the position of the sun, the weather, and the like, and changes with time of the lighting device.

  An object of the present invention includes a shape error determination device capable of appropriately determining a plurality of types of shape errors, and the shape error determination device, and as a determination result image used when presenting a shape error determination result, A determination result image generation device capable of generating an image with high visibility is provided.

In order to solve the above problems, the shape error determination device of the present invention is:
A first shape detection unit for detecting a first shape portion formed on a predetermined surface of the object from the three-dimensional shape information of the object;
A first shape abnormality determination unit that determines whether or not the first shape part is abnormal based on a detection result of the first shape detection unit;
A second shape abnormality determining unit that determines whether or not the second shape portion formed on the predetermined surface is abnormal based on the three-dimensional shape information and the detection result of the first shape detecting unit; Prepared,
The first shape abnormality determination unit compares the size of the first shape part detected by the first shape detection unit with a first shape abnormality determination threshold value, and determines whether the first shape part is abnormal. Determine whether or not
The second shape abnormality determining unit determines whether or not the second shape portion is abnormal based on three-dimensional shape information of an area where the first shape portion is not formed on the predetermined surface. It is configured.

  Accordingly, it can be determined whether or not the first shape portion is abnormal, and it can be determined whether or not the second shape portion is abnormal. Therefore, a plurality of types of shape errors are appropriately determined. be able to.

Preferably,
The first shape abnormality determination unit can be configured to set the first shape abnormality determination threshold based on a determination result of the second shape abnormality determination unit.

  By comprising in this way, the determination which considered the influence which acts on a 1st shape part and a 2nd shape part in the future becomes possible.

Also preferably,
The first shape abnormality determination unit can be configured to set the first shape abnormality determination threshold based on the size of a contact object that contacts the object.

  By comprising in this way, it can suppress that the contact object is filled in the 1st shape part, and the increase in the magnitude | size of a 1st shape part and the increase in the magnitude | size of a 2nd shape part are accelerated | stimulated. .

Also preferably,
The second shape abnormality determination unit
Generating means for generating a calculation target curve by arranging a plurality of points on the predetermined plane;
Calculation means for calculating the size of the second shape portion based on the three-dimensional shape information of the calculation target curve generated by the generation means;
A determination unit that compares the size of the second shape portion calculated by the calculation unit with a second shape abnormality determination threshold value and determines whether or not the second shape portion is abnormal;
The generating means includes
When a plurality of cross sections that are cross sections of the predetermined plane and orthogonal to a predetermined direction are extracted, and there is an area that does not include the first shape portion in the extracted cross section, an arbitrary point on the area is acquired. If there is no region that does not include the first shape portion in the extracted cross section, the calculation is performed by acquiring an arbitrary point on the cross section and arranging a plurality of points acquired from each of the extracted cross sections. Generate the target curve,
Alternatively, an edge along the predetermined direction that is an edge of the predetermined surface is extracted, a plurality of points on the extracted edge are acquired, and the calculation target curve is generated by arranging the acquired plurality of points,
The calculating means includes
When the calculation target curve generated by the generation unit does not include the first shape portion, the size of the second shape portion is calculated based on the three-dimensional shape information of the calculation target curve. When the calculation target curve generated by the generation unit includes the first shape portion, the three-dimensional shape information of the calculation target curve is corrected by interpolating the first shape portion, and the corrected calculation is performed. The size of the second shape portion can be calculated based on the three-dimensional shape information of the target curve.

  By configuring in this way, it is possible to determine whether or not the second shape portion is abnormal without excluding the first shape portion, so it is possible to appropriately determine the abnormality of the second shape portion. .

Also preferably,
The first shape portion may be a defective portion, and the second shape portion may be a worn portion.

  By comprising in this way, a defect | deletion error and a wear error can be determined appropriately.

In addition, the determination result image generation device of the present invention includes:
Comprising the shape error determination device,
It is configured to generate a determination result image in which the first shape portion and the second shape portion of the image of the object are highlighted.

  Therefore, an image with high visibility can be generated as a determination result image used when presenting the shape error determination result.

Preferably,
The first shape portion and the second shape portion may be configured to be highlighted differently.

  This configuration is preferable because the first shape portion and the second shape portion can be easily distinguished visually.

Alternatively, the determination result image generation apparatus of the present invention is
Comprising the shape error determination device,
An image of the object with the second shape portion highlighted;
An image of the object with the first shape portion enlarged;
Is generated so as to generate a determination result image.

  Therefore, an image with high visibility can be generated as a determination result image used when presenting the shape error determination result.

Preferably,
The aspect ratio of the image of the object in which the second shape portion is highlighted and the image of the object in which the first shape portion is enlarged can be made different.

  By comprising in this way, it becomes easy to grasp | ascertain visually the magnitude | size of a 2nd shape part and the magnitude | size of a 1st shape part.

According to the shape error determination device of the present invention, it is possible to appropriately determine a plurality of types of shape errors.
Moreover, according to the determination result image generation device of the present invention, an image with high visibility can be generated as a determination result image used when presenting a shape error determination result.

It is a block diagram showing the example of a structure of the shape error determination apparatus in this embodiment. It is a figure which shows an example of a target object. (A) is a figure explaining the selection method of a wear surface cross section, (b), (c) is the figure which looked at the wear surface cross section from the X direction. It is a figure which shows the example of selection of a representative wear point. It is the figure which looked at the representative wear point group from the Z direction. It is a figure explaining the detection method of a wear error. It is a figure explaining the detection method of a wear error. It is a figure explaining the setting method of a missing error judging threshold. It is a figure explaining the setting method of a missing error judging threshold. It is a block diagram showing the structural example of the determination result image generation apparatus in this embodiment. It is a figure explaining the production | generation method of a determination result image. It is a figure explaining the production | generation method of a determination result image.

  Hereinafter, embodiments of the present invention will be specifically described with reference to the drawings.

<Shape error determination device>
FIG. 1 is a block diagram illustrating a configuration example of a shape error determination device 10 according to the present embodiment. The shape error determination device 10 includes a defect detection unit 22 that functions as a first shape detection unit, a wear error determination unit 11 that functions as a second shape abnormality determination unit, and a defect error determination unit that functions as a first shape abnormality determination unit. 12 and the presence / absence of a wear error and a defect error is determined from the input three-dimensional shape information of the target object 100. In the present embodiment, as the three-dimensional shape information, three-dimensional coordinate data of a random measurement point group on the target object 100 (may be evenly spaced) is input.
Here, the “target object” is an object (for example, an object that needs to be replaced) that the user wants to determine whether there is a wear error or a defect error. “Wear error” means a state in which the amount of wear is abnormal (specifically, exceeds a predetermined wear error determination threshold), and “defect error” means that the size of the defect is abnormal. It is a state (specifically, it exceeds a predetermined missing error determination threshold).

As a method for acquiring three-dimensional shape information, for example, a method of performing stereo matching using a twin-lens camera, a method of measuring distances to a plurality of measurement points on the target object 100 using a plurality of ultrasonic distance sensors, and a laser There is a method of detecting a three-dimensional coordinate of a measurement point group using a distance meter.
For example, in the case of a method using a twin-lens camera, a relative position between two cameras (two imaging devices) is acquired in advance, and the directions of the measurement points set on the target object 100 with respect to each of the two cameras are calculated. The three-dimensional shape information can be acquired by calculating the distance from the camera to the measurement point by triangulation. At this time, all pixels in which the target object 100 in the captured image is captured may be set as measurement points to acquire the three-dimensional shape information of the entire target object 100, or a portion with a strong edge on the target object 100 may be obtained. The 3D shape information of the strong edge part may be acquired by setting only the pixel in which the image is captured as the measurement point, and the 3D shape information is information that can determine whether there is a wear error and whether there is a missing error. If it is.

The defect detection unit 22 detects a defect part (first shape part) formed on the wear surface of the target object 100 from the input three-dimensional shape information of the target object 100, and the detection result (defect detection result). Is output.
The wear error determination unit 11 determines whether or not the wear portion formed on the wear surface of the target object 100 corresponds to the wear error based on the input three-dimensional shape information of the target object 100 and the defect detection result ( Whether or not the second shape portion is abnormal) is determined, and the determination result (wear error determination result) is output.
The defect error determination unit 12 determines whether or not the defect portion formed on the wear surface of the target object 100 corresponds to the defect error based on the defect detection result and the wear error determination result (the first shape portion is abnormal). And the determination result (missing error determination result) is output.
Here, the processing of each part of the shape error determination device 10 can be realized by software processing in a CPU (Central Processing Unit) and hardware processing in an FPGA (Field-Programmable Gate Array).

FIG. 2 shows an example of the target object 100. The target object 100 is, for example, a pantograph slip plate, and the upper surface mainly becomes a contact surface (that is, a wear surface) in contact with the contact object 200 (overhead wire).
On the wear surface of the target object 100, wear due to friction with the contact object 200 exists, and in addition, defects such as a normal defect 110 and a penetration defect 120 exist. Here, the normal defect 110 is a recess that starts from one end of the wear surface and does not reach the other end, and the through defect 120 is a recess that starts from one end of the wear surface and reaches the other end. .

Hereinafter, as illustrated in FIG. 2, the longitudinal direction of the target object 100 is referred to as an X direction. Further, the vertical direction of the target object 100 is referred to as the Y direction, the upward direction is referred to as the Y positive direction, and the downward direction is referred to as the Y negative direction. The short direction of the target object 100 is referred to as the Z direction.
Note that the target object 100 is not limited to a sliding plate, and can be arbitrarily changed as appropriate. Further, the contact object 200 is not limited to the overhead wire, and can be arbitrarily changed as appropriate. Further, the surface to be a wear surface is not limited to the upper surface of the target object, and can be arbitrarily changed as appropriate.

Processing of the defect detection unit 22 will be described. In the present embodiment, the defect detection unit 22 detects the normal defect 110 and the penetration defect 120 separately.
First, an example of a method for detecting the normal defect 110 will be described. A plurality of curves corresponding to the cross section of the wear surface of the target object 100 (cross section orthogonal to the X direction) are set, and processing is performed on each curve. Hereinafter, the curve is referred to as a wear surface cross section. The wear surface section may be set freely, may be set at equal intervals from the end of the target object 100, or may be set randomly. FIG. 3A illustrates a method for selecting a wear surface section, and FIGS. 3B and 3C illustrate the wear surface section viewed from the X direction.

  The defect detection unit 22 detects the inclination in the YZ direction of the cross section of the wear surface, and detects an inclination change portion 111 where the inclination greatly changes in the middle. Then, as shown in FIG. 3 (b), the portion from the slope change point 111 to the end in the Z negative direction (downward) in the wear surface cross section is determined as the normal defect 110. To do. Thus, the normal defect 110 can be detected. Note that the inclination and the change amount threshold for detection as the inclination change point 111 are set in advance in the defect detection unit 22 and the like.

  Next, an example of a method for detecting the penetration defect 120 will be described. One representative wear point 121 is selected from each wear surface cross section. FIG. 4 shows an example of selecting the representative wear point 121. The representative wear point 121 may be set freely, may be selected at random, or may be the highest point in the Y direction. At this time, if the representative wear point 121 is set in a region that is not determined as the normal defect 110 in the cross section of the wear surface as shown in FIG. 4A, it is determined as the normal defect 110 in the calculation of the amount of wear described later. The amount of wear can also be calculated appropriately for the portion, which is preferable.

FIG. 5 is a diagram of the representative wear point 121 group viewed from the Z direction. By looking at the inclination of the representative wear point 121 group in the X and Y directions, an inclination change portion 122 where the inclination changes greatly in the middle is detected. A portion sandwiched between the two slope change points 122 and 122 and in which the position of the representative wear point 121 group is in the Y negative direction (downward) is determined as the penetration defect 120. At this time, if the distance between the slope change points 122 and 122 is larger than a certain distance, it is determined as a wear surface. The threshold value of the inclination and the amount of change for detection as the inclination change point 122 and the threshold value of the distance between the inclination change points 122 and 122 determined as the penetration defect 120 may be set together with the defect error determination threshold value.
By such a method, it is possible to detect the normal defect 110 and the penetration defect 120, respectively, and information that allows the user to determine the defect state desired by the normal defect 110 and the penetration defect 120 separately. Can be acquired.

  The example of the defect detection method in the defect detection unit 22 has been described above, but the defect detection method is not limited to this. For example, the wear surface edge at both ends in the Z direction of the wear surface is searched, and a tilt change point where the three-dimensional coordinate of the wear surface edge greatly changes in the Y negative direction (downward) is detected, and two inclination changes are detected. A portion that is sandwiched between points and in which the three-dimensional coordinates of the wear surface edge are in the negative Y direction may be determined as a defect. In addition, the defect detection unit 22 does not need to detect both the normal defect 110 and the penetration defect 120 but only needs to be able to detect the type of defect required by the user. For example, only the normal defect 110 may be detected. However, only the penetration defect 120 may be detected.

  Next, the process of the wear error determination unit 11 will be described. The wear error determination unit 11 calculates the wear amount of the target object 100 from the three-dimensional shape information and the defect detection result, and determines whether or not there is a wear error. The wear amount can be calculated by calculating the difference in the Y direction from the wear amount calculation reference surface 131 using the above-described representative wear point 121 group as the wear amount measurement target surface. In the present embodiment, as shown in FIG. 2, the wear amount calculation reference surface 131 is set based on the wear surface end portions 101 and 101 that hardly contact the contact object 200 among the wear surfaces. The setting method 131 can be arbitrarily changed as appropriate. For example, the shape data of the target object 100 is held in advance, the three-dimensional coordinates of parts such as screws and fixing holes of the target object 100 are obtained, and the wear amount calculation reference plane is moved from the part to a fixed positional relationship based on the shape data. Can also be set. In the defect detection unit 22, by setting the representative wear point 121 in a portion of the wear surface cross section that is not determined to be the normal defect 110, it is possible to appropriately calculate the wear amount even in a portion where the normal defect 110 is present.

The wear error determination unit 11 is a portion where the calculated wear amount is greater than or equal to a preset wear error determination threshold, that is, a portion where the wear surface is in the negative Y direction (downward) from the wear error determination reference surface 132 (see FIG. 2). 130 is determined as a wear error. That is, among the portions (wear portions) where the wear amount measurement target surface is in the Y negative direction (downward) from the wear amount calculation reference surface 131, the wear amount measurement target surface is in the Y negative direction from the wear error determination reference surface 132. The part at is the part corresponding to the wear error.
By calculating the wear amount and determining the wear error as described above, it is preferable that the presence / absence of a wear error can be determined in addition to the defect detection even in a portion where the defect has occurred. That is, if a missing part is excluded in a part where the wear amount is equal to or greater than the wear error determination threshold (part corresponding to a wear error), the excluded part cannot be detected as a wear error. The part 11 can be detected as a wear error.

A preferred example is shown in FIG. In the case of the normal defect 110 formed in the portion corresponding to the wear error such as the defect 141, it is excluded from the processing by the conventional method, and the portion where the defect 141 is formed is not detected as the wear error. In this method, a wear error can be detected appropriately.
Further, in the case of the penetrating defect 120 that exceeds the wear error determination reference surface 132 like the defect 142, the method of determining the wear amount by the distance from the reference is processed as exceeding the wear error determination threshold, and an unnecessary wear error is generated. It was detected. On the other hand, since the defect 142 can be detected as a through defect by using the method of the present embodiment, the processing method can be changed so that unnecessary wear errors can be prevented from being detected. For example, as shown in FIG. 7, with respect to the portion where the penetration defect 120 is generated, the wear amount measurement target surface 133 is set by linear interpolation between the slope change portions 122 and 122, and the wear amount calculation reference surface 131 is used. The difference between them may be used as the amount of wear. Thereby, the wear error determination part 11 can calculate the amount of wear also about a missing part.

In other words, the wear error determination unit 11 generates a calculation target curve (a group of representative wear points 121) by arranging a plurality of points on a predetermined surface (wear surface), and generates the three-dimensional shape information of the generated calculation target curve. Based on this, the size of the second shape portion (amount of wear) is calculated, and the calculated size of the second shape portion is compared with the second shape abnormality determination threshold value (wear error determination threshold value). It is configured to determine whether or not the two-shaped portion is abnormal (whether or not the worn portion corresponds to a wear error).
Specifically, the wear error determination unit 11 extracts a plurality of cross sections (wear surface cross sections) that are cross sections of a predetermined plane and are orthogonal to a predetermined direction (X direction in the case of the present embodiment). When there is a region that does not include one shape portion (deletion portion), an arbitrary point on the region is acquired (see FIG. 4A), and the first shape portion (deletion portion) is included in the extracted cross section. When there is no area not included, an arbitrary point on the cross section is obtained (see FIG. 4B), and a plurality of points obtained from the extracted plural cross sections (wear surface cross sections) are arranged. The calculation object curve (representative wear point 121 group) is configured to be generated.
Further, when the generated calculation target curve (representative wear point 121 group) does not include the first shape portion (defect portion), the wear error determination unit 11 is based on the three-dimensional shape information of the calculation target curve. The size (amount of wear) of the second shape portion is calculated. That is, the size (wear amount) of the second shape portion is calculated based on the three-dimensional shape information of the region where the first shape portion (defect portion) is not formed on the predetermined surface (wear surface), and the calculation result is Based on this, it is configured to determine whether or not the second shape portion is abnormal (whether or not the wear portion corresponds to a wear error).
On the other hand, when the generated calculation target curve (representative wear point 121 group) includes the first shape portion (defect portion), the first shape portion is interpolated (in the case of the present embodiment, linear interpolation). The three-dimensional shape information of the calculation target curve is corrected (see FIG. 7), and the size (amount of wear) of the second shape portion is calculated based on the corrected three-dimensional shape information of the calculation target curve. Has been. That is, the three-dimensional shape information is corrected by interpolating the first shape portion (defect portion), and the size (wear amount) of the second shape portion is calculated based on the corrected three-dimensional shape information. Based on the result, it is configured to determine whether or not the second shape portion is abnormal (whether or not the wear portion corresponds to a wear error).

Next, processing of the missing error determination unit 12 will be described. The missing error determination unit 12 determines the presence or absence of a missing error from the missing detection result. The defect detection unit 22 outputs, for example, three-dimensional coordinates of each defect part as a defect detection result.
Specifically, the missing error determination unit 12 calculates, for example, the size of each defect based on the defect detection result, and compares the calculated defect size with a preset defect error determination threshold, It is determined whether it corresponds to a missing error. The defect error determination threshold is set for the length in a single direction such as the size of the defect in the X direction, the size of the defect in the Y direction, the size of the defect in the Z direction, It is possible to set the size of the defect in the Y direction, that is, the area, or set the size of the defect in the X direction, the Y direction, and the Z direction, that is, the volume. It is also possible to set a threshold value that combines length, area, volume, and the like. The defect error determination threshold set in this way and the defect size are compared, and the defect portion detected by the defect detection unit 22 is detected as a defect state that the user wants to detect (a defect with a magnitude greater than or equal to the defect error determination threshold). ) And, if applicable, can be detected as a missing error.
In other words, the defect error determination unit 12 compares the size of the first shape portion (deletion portion) detected by the defect detection unit 22 with the first shape abnormality determination threshold value (deletion error determination threshold value). It is configured to determine whether or not one shape part is abnormal (whether or not the missing part corresponds to a missing error).

The defect error determination threshold may be set freely by the user, but since a large defect is likely to be concentrated and worn by the contact object 200 contacting the wear surface, the shape of the contact object 200 It is preferable to set according to the size. At this time, for example, when the 3D shape information of the contact object 200 can be acquired together with the 3D shape information of the target object 100, the defect error determination threshold is automatically set from the 3D shape information of the contact object 200. This is preferable because an appropriate threshold can be set.
A method for setting the defect error determination threshold according to the shape and size of the contact object 200 will be described with reference to FIG. In the example shown in FIG. 8, the contact object 200 has a cylindrical shape, the defect 151 is a normal defect 110 having a size that the contact object 200 can be filled in, and the defect 152 has a size and shape that makes the contact object 200 difficult to fill. This is a penetration defect 120. Therefore, it is preferable to set a defect error determination threshold value that determines that the defect 151 corresponds to the defect error while the defect 152 does not correspond to the defect error.

  In addition, even if the defect is the same size, a defect formed in a portion where the amount of wear is large (close to the wear error determination reference surface 132) has a high risk. Therefore, it is preferable to automatically set the defect error determination threshold according to the amount of wear. . FIG. 9 shows a missing example of the target object 100. The defect 161 is present in a more worn area, that is, an area where there are more opportunities for contact with the contact object 200 than the defect 162. Even when the defect 161 is smaller than the defect 162 or the same size, the defect 161 has more opportunities to contact the contact object 200 than the defect 162, and the contact object 200 may be caught by the defect and wear intensively. High nature. Therefore, it is preferable to change the defect error determination threshold according to the wear amount so that it is determined that even a small defect corresponds to a defect error in a region where the wear amount is large. In other words, by changing the defect error determination threshold between a defect in a region with a small amount of wear and a defect in a region with a large amount of wear, it is possible to make a determination in consideration of the effect on the defect or wear in the future.

At this time, as the defect error determination threshold value to be compared with the defect size at each wear amount, one threshold value at a specific wear amount is set, and the threshold values at other wear amounts are calculated as a ratio. This is preferable because the number of settings can be reduced. For example, a defect error determination threshold value used for a defect in a region where the wear amount is 0 is set (stored) in advance, and a defect in a region where the wear amount is not 0 decreases as the wear amount increases. Using the value obtained by multiplying the set (stored) missing error determination threshold by a coefficient of 0 to 1 to be used as the missing error determination threshold, the missing error determination threshold is changed according to the wear amount. be able to.
That is, the defect error determination unit 12 may set a first shape abnormality determination threshold value (defect error determination threshold value) based on the size of the contact object 200 that contacts the object (target object 100), or wear error. A first shape abnormality determination threshold value (defect error determination threshold value) may be set based on the determination result (wear error determination result) of the determination unit 11. Further, the first shape abnormality determination threshold value (deletion error determination) is based on both the size of the contact object 200 that contacts the object (target object 100) and the determination result (wear error determination result) of the wear error determination unit 11. Threshold) may be set.

  As described above, the defect detection and the wear amount determination are performed separately, and the presence / absence of the wear error and the presence / absence of the defect error are determined separately, thereby making it possible to perform the shape error determination of the target object in accordance with the actual situation. This is preferable.

  Here, the defect detection unit 22 of the present embodiment sets the wear surface cross section and determines the defect, but it can also be detected by other methods. For example, the outline of the wear surface of the target object 100 may be detected, and the defect may be determined from the displacement amount or shape of the outline.

In addition, the wear error determination unit 11 of this embodiment sets the representative wear point 121 group and calculates the wear amount, but it can also be calculated by other methods. For example, with respect to the wear surface edge of the target object 100, the difference from the wear amount calculation reference surface 131 in the Y-direction coordinate is used as the wear amount, and from the defect start point to the defect end point of the wear surface edge for the portion where the defect is detected. Linear interpolation may be performed, and a difference in the Y direction from the wear amount calculation reference surface 131 may be used as the wear amount based on the interpolation result.
That is, the wear error determination unit 11 extracts an edge that is an edge of a predetermined surface (wear surface) and extends along a predetermined direction (X direction in the case of the present embodiment), and acquires a plurality of points on the extracted edge. Then, a calculation target curve (wear surface edge) is generated by arranging a plurality of acquired points.
Then, when the generated calculation target curve (wear surface edge) does not include the first shape portion (deletion portion), the wear error determination unit 11 determines the first error based on the three-dimensional shape information of the calculation target curve. Calculate the size (wear amount) of the two shape parts, and if the generated calculation target curve (wear surface edge) includes the first shape part (defect part), the first shape part is interpolated (this implementation) In the case of the form, the three-dimensional shape information of the calculation target curve is corrected by linear interpolation), and the size (wear amount) of the second shape portion is determined based on the corrected three-dimensional shape information of the calculation target curve. It may be calculated.

<Determination result image generation device>
FIG. 10 is a block diagram illustrating a configuration example of the determination result image generation device 1 according to the present embodiment. The determination result image generation device 1 includes a shape error determination device 10 and an image generation unit 30, and has a wear amount, a defect detection result, a wear error determination result, a defect error determination result, and the like obtained by the shape error determination device 10. In addition, the image generation unit 30 generates a determination result image. The image generation unit 30 can be realized by software processing in a CPU, hardware processing in an FPGA, or the like.

An example of a determination result image generation method will be described with reference to FIG. FIG. 11 shows a determination result image 300 in which the missing portion and the worn portion of the image 301 of the target object 100 are highlighted.
For example, a photographed image can be used as the image 301 of the target object 100. If the method using a twin-lens camera is used for the measurement, the one of the two-lens camera may be used. When another method is used for measurement, a camera is prepared separately, and a captured image of the target object 100 is acquired by the camera simultaneously with the measurement. Note that the image 301 is not limited to a captured image, and may be a CG image or the like.

  The image generation unit 30 superimposes a wear error warning display on the image 301 of the target object 100 to emphasize a portion corresponding to the wear error. As a method of superimposing the wear error warning display, for example, a method of superimposing a mark 302 on a wear error corresponding portion of the image 301 of the target object 100 (see FIG. 11), or a wear superimposed on the wear surface edge of the image 301 of the target object 100. For example, the surface edge display 303 may be enhanced by marking by changing the color of the portion corresponding to the wear error.

For example, the wear surface edge is used for calculation when a plane including the side surface of the target object 100 (for example, a side surface corresponding to the side surface 304 in FIG. 11) is calculated and the wear error determination unit 11 calculates the wear amount. What is necessary is just to make it the position which projected the three-dimensional coordinate of the wear amount measurement object surface which has been projected on the plane containing a side surface. As another method of detecting the wear surface edge, the inclination of the wear surface cross section excluding the defect portion detected by the defect detection unit 22 is calculated, and the intersection of the extension line of the wear surface cross section and the plane including the side surface is determined as the wear surface edge. It is good. By using the inclination of the cross section of the wear surface, it is preferable that the wear surface edge display 303 can be correctly superimposed even when the wear surface is not horizontal.
Further, when the mark 302 is superimposed on the wear error corresponding part, it may be performed on the representative value of the measured wear surface.
In addition, a line 305 indicating a wear error determination threshold may be superimposed on the image 301 of the target object 100 (see FIG. 11). When the line 305 indicating the wear error determination threshold is superimposed, the degree of wear can be easily understood, and the remaining thickness until the wear error occurs can be visually grasped even in the portion where the wear error has not occurred (the portion not corresponding to the wear error). Therefore, it is preferable.

In addition, the image generation unit 30 superimposes a defect display indicating a defect part on the image 301 of the target object 100 to emphasize the defect part. Examples of the defect display include, for example, surrounding the defect part of the image 301 of the target object 100 with a circle 306 or the like (see FIG. 11), coloring the defect part, and the like. Defect display can be distinguished by changing the emphasis method between defect-free defects (defects smaller than the defect error determination threshold) and defects corresponding to defect errors (defects larger than the defect error determination threshold). It becomes easy and is suitable. For example, the drawing color is changed between a defect having no problem and a defect corresponding to a defect error, and a label 307 indicating whether the defect corresponds to a defect error is superimposed (see FIG. 11).
As described above, it is preferable that the wear error warning display and the defect display superimposed on the image 301 of the target object 100 are made different from each other so that the wear error and the defect error can be easily distinguished visually. is there.

That is, the image generation unit 30 generates the determination result image 300 in which the first shape portion (deletion portion) and the second shape portion (wear portion) of the image 301 of the object (target object 100) are highlighted. It is configured as follows. In the determination result image 300, the portion corresponding to the wear error (the portion where the wear amount is equal to or greater than the wear error determination threshold) among the wear portions is highlighted, but the wear display area 410 of the determination result image 400 described later is displayed. As described above, it is possible to highlight the entire wear portion, and to perform different highlighting on the portion corresponding to the wear error and the portion not corresponding to the wear error in the wear portion.
Further, the image generation unit 30 is configured to perform different highlighting on the first shape portion (defect portion) and the second shape portion (wear portion).

Another example of the determination result image generation method will be described with reference to FIG. FIG. 12 shows a determination result image 400 having a wear display area 410 and a defect display area 420.
In the wear display area 410, an image 411 of the target object 100 with the wear portion highlighted is superimposed. For example, the image 411 used in the wear display area 410 is subjected to a process of changing the aspect ratio in a direction (Y direction in this embodiment) in which wear occurs with respect to the image 301 (a captured image, a CG image, or the like). Get by. On the other hand, the image 421 used for the defect display area 420 described later is obtained not by the process of changing the aspect ratio but by, for example, performing a process of cutting out and expanding the defect part of the image 301. It is preferable to use an image whose aspect ratio is changed so that the wear display area 410 is stretched in the direction in which wear occurs as the image 411 because the wear degree can be easily grasped visually.

The wear surface edge display is superimposed on the image 411 to emphasize the wear surface edge. Examples of the method for emphasizing the wear surface edge include a method of superimposing marks and a method of marking by changing a color.
In addition, the wear part of the wear surface edge display changes the emphasis method so that it can be visually determined whether it is a wear part by changing the shape or color of the mark or the color of the marking. Indicates the part. At this time, the wear portion may be shown so that two types of wear states of “no problem” and “wear error” can be discriminated, but a portion where the wear amount is close to the wear error determination threshold is shown as a warning in FIG. As shown in FIG. 12, the “no problem” portion is highlighted (in the example shown in FIG. 12) or the “wear error” portion is highlighted (in the example shown in FIG. In the example shown, Δ mark) is preferable because it is possible to determine a portion where a wear error is likely to occur in the future.
In addition, a line 412 indicating a wear error determination threshold may be superimposed on the image 411 of the target object 100 (see FIG. 12). If the line 412 indicating the wear error determination threshold is superimposed, the degree of wear can be easily understood, and the remaining thickness until the wear error occurs can be visually grasped even in the portion where the wear error does not occur (the portion not corresponding to the wear error). Therefore, it is preferable.
In the wear display area 410, as shown in the determination result image 300 shown in FIG. 11, it is possible to emphasize only a portion corresponding to the wear error and not emphasize a portion not corresponding to the wear error. .

  In the defect display area 420, an image 421 obtained by cutting out a defect portion detected by the defect detection unit 22 from an image 301 (a captured image, a CG image, or the like) is used. At this time, the enlargement ratio is changed from that of the image 411 in the wear display area 410, such as enlarging the clipped image. By changing the enlargement ratio of the image of the target object 100 between the wear display area 410 and the defect display area 420, the wear surface edge display can be confirmed as a whole, and the defect can be visually confirmed in detail. Further, when the defect display area 420 changes the emphasis method between a defect having no problem (defect having a size smaller than the defect error determination threshold) and a defect corresponding to the defect error (defect having a size larger than the defect error determination threshold). Good. For example, a method of changing the color of the missing display area 420 or superimposing a label indicating whether or not a missing error has occurred.

  The defect display area 420 has an angle change defect display area 420a. From the three-dimensional coordinates of the defect portion calculated by the defect detection unit 22, the three-dimensional shape of the defect portion is known. An image 301 (a photographed image, a CG image, or the like) is pasted as a texture on the three-dimensional shape of the missing part, and an image whose angle is changed three-dimensionally is superimposed on the angle-changed defect display area 420a. It is preferable to change the angle three-dimensionally because the shape and size of the defect can be visually and intuitively easily understood.

That is, the image generation unit 30 includes an image 411 of the object (target object 100) in which the second shape portion (wear portion) is highlighted, and an object (target object 100) in which the first shape portion (deletion portion) is enlarged. It is also possible to generate a determination result image 400 including the image 421.
In this case, the image generation unit 30 includes an image 411 of the object (target object 100) in which the second shape part (wear part) is highlighted, and an object (target object 100) in which the first shape part (deletion part) is enlarged. ) And the image 421 are configured to have different aspect ratios.
The image generation unit 30 includes an image 411 of the object (target object 100) in which the second shape portion (wear portion) is highlighted, and an object (target object 100) in which the first shape portion (deletion portion) is enlarged. The aspect ratio of the image 421 may not be different.
In addition, the image generation unit 30 includes an image 411 of the object (target object 100) in which the second shape portion (wear portion) is highlighted and the first shape portion (deletion portion) is enlarged. It is not necessary to make the enlargement ratio different from the image 421 of the object (target object 100).

  Further, the defect detection unit 22 determines the normal defect 110 and the penetration defect 120. However, when emphasizing the defect part in the determination result images 300 and 400, the enhancement method is changed between the normal defect 110 and the penetration defect 120. May be. By changing the emphasis method between the normal defect 110 and the through defect 120, for example, since the contact object 200 is likely to be filled, the defect error at the through defect 120 is important. This is preferable because it is easy to visually determine when the importance changes.

In addition, in the determination result image 400, when there are a plurality of missing portions, if the missing display area 420 is set for all the missing portions, each enlarged image (each image 421) becomes smaller. It is good to change the priority to do. For example, it is set so that an image obtained by cutting out a defective portion corresponding to a high-priority defective error is superimposed and an image obtained by cutting out another defective portion is not superimposed. At this time, the user can confirm all the missing portions by switching the image obtained by cutting out other missing portions or by switching the images in terms of time. Moreover, the enlargement rate may be made larger than other defects, assuming that the defect most deviating from the missing error determination threshold is the degree of the defect error.
As described above, by determining wear errors and missing errors separately and generating a determination result image according to the results, it is possible to present errors in an appropriate shape that is easier to understand visually. it can.

  According to the shape error determination apparatus 10 of the present embodiment described above, the first shape portion (defect portion) formed on the predetermined surface (wear surface) of the object (target object 100) from the three-dimensional shape information of the object (target object 100). Whether or not the first shape portion is abnormal based on the first shape detection unit (deletion detection unit 22) for detecting the detection result and the detection result (deletion detection result) of the first shape detection unit (deletion detection unit 22). Detection result of first shape abnormality determination unit (deletion error determination unit 12) for determining (whether or not the defect part corresponds to a defect error), three-dimensional shape information, and first shape detection unit (defect detection unit 22) A second shape for determining whether or not the second shape portion formed on the predetermined surface (wear surface) is abnormal (whether the wear portion corresponds to a wear error) based on (defect detection result) An abnormality determination unit (wear error determination unit 11), and a first shape abnormality The fixed unit (deletion error determination unit 12) includes a size of the first shape portion (deletion portion) detected by the first shape detection unit (deletion detection unit 22) and a first shape abnormality determination threshold (deletion error determination threshold). To determine whether or not the first shape portion is abnormal (whether or not the defective portion corresponds to a defect error), and determine a second shape abnormality determination unit (wear error determination unit 11). Is based on the three-dimensional shape information of the region where the first shape portion (defect portion) is not formed on the predetermined surface (wear surface), whether or not the second shape portion is abnormal (the wear portion becomes a wear error). It is configured to determine whether or not this is the case.

  Therefore, it can be determined whether or not the first shape portion is abnormal (whether or not the defect portion corresponds to a defect error) and whether or not the second shape portion is abnormal (the wear portion is worn). It is possible to determine whether or not it corresponds to an error, and therefore it is possible to appropriately determine a plurality of types of shape errors.

  For example, the second shape abnormality determination unit (wear error determination unit 11) generates a calculation target curve (representative wear point 121 group or wear surface edge) by arranging a plurality of points on a predetermined surface (wear surface). Means for calculating the size (wear amount) of the second shape portion based on the three-dimensional shape information of the calculation target curve (the representative wear point 121 group or the wear surface edge) generated by the generation means; The size of the second shape portion calculated by the calculation means is compared with the second shape abnormality determination threshold value (wear error determination threshold value) to determine whether or not the second shape portion is abnormal (the wear portion is Determining means for determining whether or not the wear error is satisfied, and the generating means is a cross section of the predetermined surface and a cross section (wear surface cross section) orthogonal to a predetermined direction (X direction in the present embodiment) ) If there is an area that does not include the first shape part (deletion part) in the cut cross section, an arbitrary point on the area is acquired and the extracted cross section does not include the first shape part (deletion part) If there is not, the calculation target curve (representative wear point 121 group) is obtained by arranging a plurality of points acquired from each of a plurality of extracted cross sections (wear surface cross sections) while acquiring an arbitrary point on the cross section. Generate or extract edges along a predetermined direction (in the present embodiment, the X direction) that are edges of a predetermined surface (wear surface), and acquire a plurality of points on the extracted edge. The calculation target curve (wear surface edge) is generated by arranging the points, and the calculation means adds the first shape portion (defect portion) to the calculation target curve (representative wear point 121 group or wear surface edge) generated by the generation means. ) Is not included Calculates the size (wear amount) of the second shape portion based on the three-dimensional shape information of the calculation target curve, while calculating the calculation target curve (representative wear point 121 group or wear surface edge) generated by the generation means. ) Includes a first shape portion (a missing portion), the three-dimensional shape information of the calculation target curve is corrected by interpolating the first shape portion (in the case of the present embodiment, linear interpolation), The size (wear amount) of the second shape portion can be calculated on the basis of the corrected three-dimensional shape information of the calculation target curve.

  By configuring in this way, it is possible to determine whether or not the second shape portion is abnormal (whether or not the wear portion corresponds to a wear error) without excluding the first shape portion (defect portion). Since it can do, abnormality of the 2nd shape part (wear part) can be judged appropriately.

  Moreover, according to the shape error determination device 10 of the present embodiment, the first shape abnormality determination unit (deletion error determination unit 12) determines the determination result (wear error determination) of the second shape abnormality determination unit (wear error determination unit 11). The first shape abnormality determination threshold value (defect error determination threshold value) can be set based on the result.

  By comprising in this way, the determination which considered the influence which acts on a 1st shape part (deletion part) and a 2nd shape part (wear part) in the future becomes possible.

  Further, according to the shape error determination device 10 of the present embodiment, the first shape abnormality determination unit (deletion error determination unit 12) is based on the size of the contact object 200 in contact with the object (target object 100). It is possible to configure so as to set one shape abnormality determination threshold (missing error determination threshold).

  With this configuration, the first shape portion (defect portion) is filled with the contact object 200, and the size of the first shape portion (deletion portion) increases or the size of the second shape portion (wear portion). It is possible to prevent the increase in the amount from being promoted.

  In the present embodiment, the first shape portion is a defect portion and the second shape portion is a wear portion. However, the first shape portion is not limited to the defect portion, and the second shape portion is not limited to the wear portion. For example, the first shape portion may be a portion concealed by a foreign object. Further, the second shape portion may be a curved portion or the like.

  According to the determination result image generation device 1 of the present embodiment described above, the shape error determination device 10 is provided, and the first shape portion (defect portion) and the second shape portion of the image 301 of the object (target object 100). The determination result image 300 in which (wear part) is highlighted is generated.

  Therefore, an image with high visibility can be generated as the determination result image 300 used when presenting the shape error determination result.

  In addition, according to the determination result image generation device 1 of the present embodiment, it is possible to configure so that different highlights are displayed in the first shape portion (defect portion) and the second shape portion (wear portion). is there.

  This configuration is preferable because the first shape portion (defect portion) and the second shape portion (wear portion) can be easily distinguished visually.

  Alternatively, according to the determination result image generation device 1 of the present embodiment described above, the image 411 of the object (target object 100) in which the second shape portion (wear portion) is highlighted and the first shape portion (defect portion). ) And an image 421 of the enlarged object (target object 100).

  Therefore, an image with high visibility can be generated as the determination result image 400 used when presenting the shape error determination result.

  Further, according to the determination result image generation device 1 of the present embodiment, the image 411 of the object (target object 100) in which the second shape portion (wear portion) is highlighted and the first shape portion (deletion portion) are enlarged. The aspect ratio of the image 421 of the selected object (target object 100) can be made different.

  By comprising in this way, it becomes easy to grasp | ascertain visually the magnitude | size of a 2nd shape part (wear part) and the magnitude | size of a 1st shape part (deletion part).

  In each of the above-described embodiments, the configuration and the like illustrated in the accompanying drawings are merely examples, and are not limited thereto, and can be appropriately changed within the scope of the effects of the present invention. is there. In addition, various modifications can be made without departing from the scope of the object of the present invention.

DESCRIPTION OF SYMBOLS 1 Determination result image generation apparatus 10 Shape error determination apparatus 11 Wear error determination part (2nd shape abnormality determination part)
12 Missing error determination unit (first shape abnormality determination unit)
22 Defect detection unit (first shape detection unit)
100 Target object (object)
200 Contact object 300, 400 Determination result image 301, 411, 421 Object image

Claims (9)

A first shape detection unit for detecting a first shape portion formed on a predetermined surface of the object from the three-dimensional shape information of the object;
A first shape abnormality determination unit that determines whether or not the first shape part is abnormal based on a detection result of the first shape detection unit;
A second shape abnormality determining unit that determines whether or not the second shape portion formed on the predetermined surface is abnormal based on the three-dimensional shape information and the detection result of the first shape detecting unit; Prepared,
The first shape abnormality determination unit compares the size of the first shape part detected by the first shape detection unit with a first shape abnormality determination threshold value, and determines whether the first shape part is abnormal. Determine whether or not
The second shape abnormality determining unit determines whether or not the second shape portion is abnormal based on three-dimensional shape information of an area where the first shape portion is not formed on the predetermined surface. Characteristic shape error determination device.   The shape error determination apparatus according to claim 1, wherein the first shape abnormality determination unit sets the first shape abnormality determination threshold based on a determination result of the second shape abnormality determination unit.   The shape error determination device according to claim 1, wherein the first shape abnormality determination unit sets the first shape abnormality determination threshold based on a size of a contact object that contacts the object. . The second shape abnormality determination unit
Generating means for generating a calculation target curve by arranging a plurality of points on the predetermined plane;
Calculation means for calculating the size of the second shape portion based on the three-dimensional shape information of the calculation target curve generated by the generation means;
A determination unit that compares the size of the second shape portion calculated by the calculation unit with a second shape abnormality determination threshold value and determines whether or not the second shape portion is abnormal;
The generating means includes
When a plurality of cross sections that are cross sections of the predetermined plane and orthogonal to a predetermined direction are extracted, and there is an area that does not include the first shape portion in the extracted cross section, an arbitrary point on the area is acquired. If there is no region that does not include the first shape portion in the extracted cross section, the calculation is performed by acquiring an arbitrary point on the cross section and arranging a plurality of points acquired from each of the extracted cross sections. Generate the target curve,
Alternatively, an edge along the predetermined direction that is an edge of the predetermined surface is extracted, a plurality of points on the extracted edge are acquired, and the calculation target curve is generated by arranging the acquired plurality of points,
The calculating means includes
When the calculation target curve generated by the generation unit does not include the first shape portion, the size of the second shape portion is calculated based on the three-dimensional shape information of the calculation target curve. When the calculation target curve generated by the generation unit includes the first shape portion, the three-dimensional shape information of the calculation target curve is corrected by interpolating the first shape portion, and the corrected calculation is performed. The shape error determination device according to claim 1, wherein the size of the second shape portion is calculated based on three-dimensional shape information of the target curve.   5. The shape error determination device according to claim 1, wherein the first shape portion is a defective portion, and the second shape portion is a wear portion. 6. Comprising the shape error determination device according to any one of claims 1 to 5,
A determination result image generating apparatus that generates a determination result image in which the first shape portion and the second shape portion of the image of the object are highlighted.   The determination result image generation apparatus according to claim 6, wherein different highlighting is performed for the first shape portion and the second shape portion. Comprising the shape error determination device according to any one of claims 1 to 5,
An image of the object with the second shape portion highlighted;
An image of the object with the first shape portion enlarged;
A determination result image generating apparatus characterized by generating a determination result image including   The determination result according to claim 8, wherein the aspect ratio of the image of the object in which the second shape portion is highlighted and the image of the object in which the first shape portion is enlarged are different from each other. Image generation device.

Images