JPH0953923A - Device for inspecting three-dimensional object for shape - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make the minimum size of defects to be inspected variable in accordance with the size of the defect of an object to be inspected. SOLUTION: The device for inspecting three-dimensional object for shape is provided with an inspection picture inputting section 10, a picture processing section 12 which discriminates the presence/absence of the defect of a three- dimensional object by processing inspection pictures from the section 10, and an inspection range control section 16 which outputs a standard model specifying signal to the section 12 in accordance with an object identification number for identifying object to be inspected inputted from the outside and, at the same time, the inspection range information when the image of a standard model picture is picked up to a light source section and image pickup section as inspection range signals. The section 10 is provided with the light source section equipped with a zoom mechanism for light source which reduces optical cutting rays based on a prescribed inspection range signal and a CCD camera equipped with zoom mechanism which enlarges the image of the surface of the object to be inspected based on the prescribed inspection range signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a shape inspection apparatus for inspecting the presence or absence of a defect in a standard shape by image processing, and in particular, it can change the minimum defect size of a three-dimensional object such as a sand mold to be inspected. Shape inspection device.

[0002]

2. Description of the Related Art Conventionally, in the process of manufacturing a casting by sand mold casting, when the mold is removed, the sand mold may not work well and a part may be chipped to cause a defect in the sand mold. Since the casting made by the sand mold having this defect is a defective product, it is necessary to automatically identify the defect of the mold before pouring the molten metal.

Conventionally, visual detection has been used to detect a sand-shaped defect having a three-dimensionally complicated shape in which the entire image is apparently black and has almost no contrast. In addition, there is no known case in which a defective portion due to a chip on the sand surface is detected by image processing. Therefore, we have developed a device that automatically detects defects on the surface of the sand mold during the manufacturing process by image processing, and has already filed a part of it (Japanese Patent Application No. 6-1979).
23, Japanese Patent Application No. 6-261472).

[0004]

However, in the above-mentioned conventional example, the size of the defect that can be detected is determined by the number of slits, because the entire subject is imaged as one image and is the object of inspection. However, since the number of slits that can be mounted on the projector is limited to one, the smallest detectable defect is fixed, and this is not suitable for continuous inspection of objects with different minimum defect sizes that require detection. It was Furthermore, there is an inconvenience that the detected defects cannot be examined in detail.

To solve this problem, if the number of slit plates to be mounted is set to be as large as possible, the inspection range is not narrowed due to the resolution of the camera, and the image processing time is increased. , The problem arises.

[0006]

The object of the present invention is to improve the disadvantages of the conventional example, and in particular, to make the minimum defect size of the defect to be inspected variable according to the size of the defect of the three-dimensional object to be inspected. It is an object of the present invention to provide a shape inspection device for a three-dimensional object.

[0007]

Therefore, in the present invention, as a first means, an inspection image input section for irradiating a three-dimensional object to be inspected with a light cutting line and imaging the surface of the three-dimensional object,
Extracts the mismatched portion between the inspection target image output from the inspection image input unit and the standard model image specified based on the predetermined standard model specifying signal, and also detects the defect of the three-dimensional object to be inspected based on the mismatched portion data. An image processing unit that determines the presence or absence of the three-dimensional object, and a model image storage unit that stores a plurality of standard model images obtained by previously capturing images of the three-dimensional object to be inspected for each type. Moreover, the inspection image input unit has a light source unit having a light source zoom mechanism for reducing the light section line based on a predetermined inspection range signal, and imaging for enlarging the image of the surface to be inspected based on the predetermined inspection range signal. And an image pickup unit having a zoom mechanism. Furthermore, in the inspection image input section,
The standard model specific signal is output to the image processing unit according to the inspection target identification number for external inspection target identification and the inspection range information when the standard model image is captured is used as the inspection range signal and the light source unit and the image capturing. An inspection range control unit that outputs to the department was also installed.

In the first means specified by these matters, the inspection image input unit first irradiates a complex three-dimensional object, such as a sand mold, to be inspected with a light cutting line (slit light), and further, An inspection image (inspection image data) is generated by imaging the surface of the three-dimensional object onto which the slit light is projected.

At this time, the inspection range control unit outputs a standard model specifying signal to the image processing unit according to the inspection target identification number for externally inputting the inspection target identification. Then, the standard model image of the sand type to be inspected is read from the model image storage unit. Further, the inspection range when the standard model image is captured is output to the light source section and the image capturing section of the inspection image input section as an inspection range signal.

In the light source section, the light source zoom mechanism operates in accordance with the inspection range signal to change the range in which slit light is emitted. Further, in the image pickup unit, the image pickup zoom mechanism operates in accordance with the inspection range signal to change the magnification of image pickup. Therefore, the inspection object range is variable according to the inspection object identification number input from the outside. This external input is
It may be performed by the user of this shape inspection apparatus, or a means for automatically identifying the type of the subject may be added.

The image processing unit generates the mismatched portion data by extracting the mismatched portion between the inspection target image and the standard model image. The extraction of the non-coincidence portion may be performed after binarizing the both with a constant threshold value, or may be compared without changing the gradation data. Furthermore, the image processing unit determines the presence or absence of a defect of the three-dimensional object to be inspected based on the mismatched portion data. In this case, a portion having a certain size or more in the disagreement portion may be omitted, or the closed curve portion may be omitted when the closed curve is extracted.

The first means carries out continuous processing in order to change the inspection range by external input even when there are a plurality of types of specimens conveyed to the inspection position.

In the second means, in addition to the matters for specifying the first means, when the inspection range control section outputs an inspection range signal corresponding to the inspection object identification number to the light source section, the size of the inspection range is determined. A light amount control function for controlling the illuminance of the light source unit is provided.

When the irradiation range of the slit light is changed, the second means changes the luminous intensity of the light source accordingly, so that the brightness of the surface of the three-dimensional object to be inspected becomes constant. There is. By keeping the brightness constant, the extraction processing of the part that does not match the model image and the binarization accuracy are stabilized.

In the third means, in addition to the matter for specifying the first or the second means, the image processing section causes the defect generation position and the defect size in the three-dimensional object to be inspected on the basis of the mismatched data. The inspection range control unit has a detailed inspection control function of outputting an inspection range signal to the light source unit and the imaging unit based on the defect occurrence position information calculated by the defect part calculation function. It was

The third means performs a more detailed inspection according to the inspection result obtained by the first or second means. The defective portion calculation function calculates the defect occurrence position based on the number of pixels and the coordinates of the mismatched portion. The detailed inspection control function enlarges the portion based on the defect occurrence position information and the size information. The defect is inspected again based on the inspection image enlarged by the detailed inspection control function. At this time,
Since the number of slit light beams applied to the defective portion is larger than that in a normal inspection, more precise inspection is performed.

The enlargement ratio may be fixed at a plurality of enlargement ratios, or the enlargement ratio may be determined based on the defect size information calculated by the defect portion calculation function. In this case, the standard model image is enlarged or reduced according to the enlargement ratio. Regarding the positional relationship between the subject and the inspection image input unit, either one may be moved.

In the fourth means, in addition to the matter for specifying the first or second means, the inspection image input section supports the imaging section and the light source section, and the imaging section and the light source are based on a predetermined positioning signal. There is also a positioning unit that moves the unit horizontally. In addition, the inspection range control unit specifies a division inspection region in which the defect has occurred based on the defect occurrence position information among the division inspection regions obtained by dividing the inspection range in advance, and the division inspection region specifying function. And a positioning function for outputting a positioning signal to the positioning unit based on the position information of the divided areas specified by the inspection area specifying function.

Like the third means, the fourth means carries out a more detailed inspection according to the inspection result by the first or second means. Here, the inspection range is divided into a plurality of parts for a detailed inspection in advance, and the detailed inspection is performed based on this divided area.

First, the divided inspection area specifying function determines which divided area the defective divided area is. Next, the positioning function outputs a positioning signal to the positioning unit so that the defective divided area is directly below the inspection image input unit. Then, the positioning unit moves the imaging unit and the light source unit according to the positioning signal. Further, the inspection range control unit outputs an inspection range signal to the inspection image input unit so that the divided area becomes the inspection range. Further, in the fourth means, the model image storage unit stores standard model images in the range of each divided area.

The present invention aims to achieve the above-mentioned object by these means.

[0022]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described below with reference to the drawings.

1 and 2 are block diagrams showing the configuration of the shape inspection apparatus for a three-dimensional object according to the present invention. The shape inspection apparatus has an inspection image input unit including a light source unit (projector) 22 that irradiates a three-dimensional object to be inspected with a light cutting line (slit light), and a CCD camera (imaging unit) 18 that images the surface of the three-dimensional object. Equipped with 10.

Moreover, the shape inspection apparatus includes an image processing unit 12 for extracting a mismatched portion between the inspection target image output from the inspection image input unit 10 and the standard model image specified based on a predetermined standard model specifying signal. A model image storage unit 14 that stores a plurality of standard model images captured in advance for each type of three-dimensional object to be inspected, and an inspection range control unit 16 that controls the inspection range are provided.

As shown in FIG. 2, the inspection range control unit 16
Outputs a standard model specifying signal to the image processing unit 12 according to the inspection object identification number for inspection object identification input from the outside, and an inspection range when the standard model image is captured as an inspection range signal b. Unit 22 and imaging unit 18
And the light amount control function 16B for controlling the illuminance of the light source unit according to the size of the inspection range 3 when outputting the inspection range signal according to the inspection target identification number to the light source unit 22. It has and. The inspection range control unit 16 is actually realized by the operation of the host computer connected to the inspection image input unit 10.

The light source unit 22 is provided with a light source zoom mechanism 24 for reducing the light section line based on a predetermined inspection range signal, while the CCD camera 18 is an image of the surface to be inspected based on the predetermined inspection range signal. The imaging zoom mechanism 20 for enlarging the image is provided.

As shown in FIG. 3A, the light source unit 22 is
A lamp 30, a mirror 31 for reflecting the light emitted from the lamp in the sand mold 1 direction, a lens 32 for condensing the light from the lamp 30 and the mirror 31 on a slit plate 34, and a slit plate 34 are provided. ing. Further, the light source unit 22 includes a light source zoom mechanism 24 that makes the irradiation range of the light cutting line (slit light) from the slit plate variable, and the light source zoom mechanism 24 includes a zoom lens 36.

As shown in FIG. 3A, the slit light from the light source section 22 having such a configuration is projected onto the sand mold 1 as shown in the figure. FIG. 3 (B) is a partially enlarged view of a standard model having no defect, and FIG. 3 (C) is a partially enlarged view of an inspection image of a sand mold having a defect. The standard model image is captured in advance and the model image storage unit 14
It is stored in.

When the inconsistency between this inspection image and the standard model image is extracted, the image shown in FIG. 3D is obtained. Then, the slit light corresponding to the defective portion 2 appears as a closed curve by taking the logical difference of the images. Therefore, if a logical difference between the inspection image and the standard model image is taken and a closed curve is further extracted, this is regarded as the defective portion of the sand mold 1.

In the processing by this closed curve extraction, even if noise occurs during image processing, it is difficult for the noise to become a closed curve. Therefore, the processing in which the noise is removed by making the portion where the closed curve occurs a missing portion is performed. be able to.

Further, instead of this closed curve extraction processing, the inspection image and the standard model image are extracted as unmatched portions with the data having gradation, and further, the portion where the gradation difference is larger than a certain threshold value is extracted. Labeling may be performed to determine a label portion having a certain size or more as a defective portion.

FIG. 4 is a front view showing the structure of the inspection image input section 10. In this embodiment, the light source unit (projector)
A light source zoom mechanism 24 and a light amount adjustment mechanism 23 (not shown) are provided at 22, and an imaging zoom mechanism 20 is also provided at the CCD camera 18. As shown in FIG. 4, the sand mold 1 in the production process is conveyed to the inspection position by the belt conveyor 2. Since there are various sizes of the sand mold 1 conveyed to this inspection position, the inspection range 3 changes depending on the type of sand mold.

FIG. 5 shows an example in which a sand mold having an inspection range 3 smaller than that shown in FIG. 4 is targeted. FIG.
In the example shown in FIG. 4, the irradiation mechanism of the slit light is reduced by operating the light source zoom mechanism 24, and the magnification of the CCD camera is increased by operating the imaging zoom mechanism 20. I am raising. Therefore, the range that can be inspected and the minimum defect that can be detected can be made variable. Therefore, even if the type of the object that is successively transported to the inspection position changes, the inspection range can be changed according to this change. You can

The inspection range control section 16 outputs a standard model specifying signal to the image processing section 12 in accordance with the inspection object identification number for externally inputting the inspection object identification, and also in the specified standard model image. Inspection range information is the inspection range signal b
Is output to the light source unit 22 and the imaging unit 16. The inspection target identification number is information for identifying the sand mold model number (type) here, and the inspection range control unit 16 specifies a standard model image used for image processing based on the inspection target identification number, Further, the inspection range is determined. The inspection range signal b is actually a control signal for driving and controlling the light source zoom mechanism 24 and the imaging zoom mechanism 20.

When the projection area of the slit light is reduced by the light source zoom mechanism 24, the projection surface becomes bright accordingly. If the captured image does not have a constant brightness, the accuracy of image processing changes, so here the illuminance of the light source is adjusted according to the change in the enlargement ratio so that the brightness of the captured image is always constant. . In the example shown in FIG. 4, the light amount of the projector 22 is adjusted. As another method, the diaphragm of the CCD camera 18 may be adjusted or a filter may be used.

FIG. 6 is an explanatory view showing an example of a standard model, FIG. 6 (A) shows before slit light projection, and FIG. 6 (B).
Indicates a state in which slit light is projected. On the other hand, FIG.
Shows an example of a sand mold in which a defect has occurred, and a defect has occurred in the upper left part from the center of FIG. 7 (A). When the image after slitting is projected on this, it becomes as shown in FIG. 7 (B). When the standard model image shown in FIG. 6 (B) and the inspection image shown in FIG. 7 (B) are binarized and then a logical difference is taken, and only the closed curve is extracted, the defective portion becomes as shown in FIG. Extracted well. In addition, the standard model image shown in FIG.
Even if the difference between the gradation values is calculated for each pixel with the image data having gradation as compared with the inspection image shown in FIG. good.

The image processing section 12 calculates the position and size of the sand-shaped defect portion based on the coordinates of the closed curve extracted in this way. This is the number of pixels and each zoom mechanism 2
The length in the real space per dot of the inspection image may be obtained from the enlargement ratios used for 0 and 24, and may be obtained from the number of pixels forming the closed curve.

[Detailed Inspection] Next, a method for inspecting the details of the defect after detecting the defect by the above method will be described.

In the method of performing the detailed inspection processing, the image processing unit 12 calculates the defect generation position and the defect size generated in the three-dimensional object to be inspected based on the closed curve extracted from the mismatched portion extraction image. It has a partial calculation function. Moreover, the inspection range control unit 16 has a detailed inspection control function of outputting the inspection range signal b to the light source unit 22 and the CCD camera 18 based on the defect occurrence position information calculated by the defect portion calculation function.

The detailed inspection control function is a function for controlling the operation of the inspection image input unit 10 based on the position and size to be zoomed up in order to zoom up the defective portion. First, the defect portion calculation function calculates the defect occurrence position based on the number of pixels and the coordinates of the mismatched portion.
Next, the detailed inspection control function enlarges the portion based on the defect occurrence position information and the size information. The defect is inspected again based on the inspection image enlarged by the detailed inspection control function. At this time, since the number of slit lights radiated to the defective portion is larger than that in the normal inspection, more precise inspection is performed.

Regarding the determination of the enlargement ratio by the detailed inspection control function, the enlargement ratio may be fixed and selected at a plurality of enlargement ratios, or enlargement may be performed based on the defect size information calculated by the defect portion calculation function. The rate may be determined. In this case, the standard model image is enlarged or reduced according to the enlargement ratio. Regarding the positional relationship between the subject and the inspection image input unit, the imaging angle of the camera 18 or the like may be changed or the subject may be moved.

An example in which an inspection image is generated by enlarging at a predetermined enlargement ratio and a predetermined center position will be described. Here, the position at which the inspection region is divided into a plurality of regions is defined for each type of sand mold of the subject.

In the example of performing the detailed inspection based on the divided areas, the inspection image input unit 10 supports the CCD camera 18 and the light source unit 22, and the CCD camera 18 and the light source unit 22 are supported based on a predetermined positioning signal. A positioning unit 28 for moving horizontally is also provided (see FIG. 4). The positioning unit 28 includes a driving unit 28A and a guide 28B.

Further, the inspection range control unit 16 specifies the divided inspection area in which the defect has occurred based on the defect occurrence position information among the divided inspection areas in which the inspection range is previously divided into a plurality of areas. It has a function and a positioning function for outputting a positioning signal to the positioning unit based on the position information of the divided area specified by the divided inspection area specifying function.

FIG. 9 shows an example of the definition of the divided areas. If FIG. 9A is taken as the normal inspection range, FIG.
The divided areas divided into four in FIG. 9B and eight in FIG. 9C are the inspection range of the detailed inspection. The model image storage unit 14 stores standard model image data for each divided area at each enlargement ratio shown in FIG.

FIG. 10 is a table showing an example of control contents by the inspection range control unit 16. The inspection range control unit 16 specifies the type of the sand mold to be inspected based on the inspection object identification number from the outside, further specifies the standard model image, and the inspection range and the light amount when the standard model image is captured. Control to reproduce.

For this reason, the inspection range control unit 16 stores the zoom magnification and the light amount corresponding to the type of sand mold as control information. In FIG. 10, instead of the actual control value,
For the sake of explanation, an example is shown in which the control values for the normal inspection range of the sand mold 1 are set to "1", respectively.

When the inspection object identification number is input, for example, "sand mold 2" is specified according to this number. Further, when a defect is found in the normal inspection range and a detailed inspection is performed, the magnifications of the imaging zoom mechanism 20 and the light source zoom mechanism 24 are first determined according to the control table shown in FIG. Then, the inspection range signal b is sent to each zoom mechanism 2
Output to 0 and 24.

Regarding the brightness of the light source, the projector 2
Since the amount of light is changed by changing the voltage supplied to No. 2, the voltage defined for each type of sand mold is supplied to the projector 22. In the example shown in FIG. 10, the value of the light amount is the value of the light amount output from the projector (light source unit) 22, and is not the supply voltage value. The supply voltage value is determined by the type of projector and the like.

The inspection range control unit 16 specifies the divided area based on the position information on the inspection image where the defect is found. For example, it is specified as the lower left area of FIG. The inspection range control unit 16 outputs the standard model specifying signal a corresponding to the specified divided area, and the positioning unit 28 outputs the positioning signal so that the lower left area is directly below the light source section 22.
Output to

In this way, it is possible to satisfactorily obtain an inspection image for detailed inspection in which the density of slit light is increased,
Therefore, even if the shape of the sand mold is fine and complicated, the shape inspection can be satisfactorily performed.

[Overall Operation] Next, the overall operation will be described with reference to the flowcharts of FIGS. 11 and 12.

First, when the sand mold, which is the subject, is transported to the inspection position, the inspection object identification number is input (step S
1). This inspection object identification number may be input by the operator, or the sand type positioning hole 40 (see FIG. 9).
The inspection target may be recognized based on the diameter of the reference) and the inspection target identification number of the recognized type may be input to the inspection range control unit 16.

Further, the inspection range control unit 16 controls the magnification of the light source zoom mechanism 24 and the light amount of the light source unit 22 according to the inspection object identification number (step S2). This is performed by outputting the inspection range signal b and the light amount control signal c according to the control information shown in FIG.

Next, the inspection range control unit 16 controls the magnification of the image pickup zoom mechanism 20 of the CCD camera 22 by the inspection range signal b. (Step S3). With the inspection range and the light amount fixed, the inspection image input unit 10 images the surface of the sand mold on which the slit light is projected and inputs the inspection image to the image processing unit 12 (step S4).

The inspection range control unit 16 outputs the standard model specifying signal a to the image processing unit 12 according to the inspection object identification number, and the image processing unit 12 generates the standard model image based on the standard model specifying signal. It is read from the model image storage unit 14 (step S5).

The image processing section compares the inspection image with the standard model image to extract the defective portion (step S6).
This may be due to a closed curve or due to a gradation difference.

If no defect is detected (step S
7), "OK" is output. Then, the next sand mold is transported (step S8), and the process returns to step S1.

On the other hand, when a defect is detected (step S
7), "NG" is output, and it is further confirmed whether or not a detailed inspection is to be performed (step S9). Conditions for detailed inspection,
For example, when detailed inspection is performed depending on the type of sand mold and whether detailed inspection is not performed, or when detailed inspection is performed according to conditions such as the size of a detected defect, the details shown in FIG. Move to inspection processing.

If the detailed inspection is not performed, the inspection result output processing such as the position and size of the defect is performed (step S).
10). Further, the next subject is transported, and (step S1
1), the process returns to step S1.

In the detailed inspection processing, the inspection range control unit 16
However, the divided area including the position where the defect occurs is specified based on the position information of the defective portion detected in step S7 (step S21). When the defective portion extends over two areas, the divided areas are inspected twice.

Further, the inspection range control unit 16 projects the slit light on the specified divided area and picks up the image, so that the positioning unit 28 outputs the positioning signal according to the specified divided area.
Then, the positioning unit 28 moves the CCD camera 18 and the light source 22 (step S22).

Further, the inspection range control unit 16 controls the zoom mechanisms 20 and 24 and the light amount (steps S23-).
24), the CCD camera 18 images the divided area onto which the slit light with high density is projected (step S25).

Further, the image processing unit 12 inspects the defect based on the standard model image of the specified divided area and the inspection image picked up in step S25 (step S25).
26). Next, the details of the inspection result obtained by this image processing are output to the outside (step S27). As a result, the inspection result of the inspection having the smaller minimum defect size is output. Further, the next subject is transported (step S28), and the process returns to step S1.

The processing shown in FIGS. 11 to 12 can cover a wide inspection range and a narrow inspection range with one set of inspection system, and thus the versatility of the shape inspection apparatus can be enhanced.

[0066]

Since the present invention is constructed and functions as described above, according to this, the inspection range control unit changes the inspection range in accordance with the inspection object identification number for the inspection object identification inputted from the outside. Since the magnification of the imaging zoom mechanism and the light source zoom mechanism is controlled, the inspection range can be widely varied, and therefore stable continuous inspection is possible without being affected by the size of the subject. Since only the defective portion can be enlarged and inspected, more detailed inspection data can be obtained. As described above, it is possible to provide an unprecedented excellent shape inspection device for a three-dimensional object, in which the minimum defect size of the defect to be inspected can be changed according to the size of the defect of the three-dimensional object to be inspected.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of the present invention.

FIG. 2 is a block diagram showing configurations of an inspection image input unit and an inspection range control unit shown in FIG.

3A and 3B are views for explaining a processing example of the image processing unit shown in FIG. 1, FIG. 3A is a view showing a state in which an inspection image is captured, and FIG. 3B is a standard model image. In the figure showing an example,
FIG. 3C is a diagram showing an inspection image of a sand mold with a defect,
FIG. 3D is a diagram illustrating an example of the defect portion extraction processing.

FIG. 4 is a front view showing the positional relationship of the inspection image input unit shown in FIG.

5 is a front view showing a case of an inspection range different from the inspection range shown in FIG.

6 is an explanatory diagram showing an example of a standard model image, and FIG.
(A) is a figure which shows the image which imaged the sand mold surface before slit light projection, and FIG.6 (B) is a figure which shows the image which imaged the sand mold surface after slit light projection.

FIG. 7 is an explanatory diagram showing an example of an inspection image obtained by imaging a defective sand mold, FIG. 7 (A) is a diagram showing an image of the sand mold surface before slit light projection, and FIG. 7 (B) is a slit; It is a figure which shows the image which imaged the sand mold surface after light projection.

FIG. 8 is an explanatory diagram showing an example in which a closed curve is extracted from the images shown in FIGS. 6B and 7B.

9A and 9B are explanatory diagrams showing an example of division of an inspection range when a detailed inspection is performed by the configuration shown in FIG. 1, FIG. 9A is a diagram showing a normal inspection range, and FIG. FIG. 9C is a diagram showing an example in the case of being divided into three, and FIG. 9C is a diagram showing an example in the case of being divided into eight.

10 is a table showing an example of control information used in the inspection range control unit shown in FIG.

11 is a flowchart showing processing steps in the configuration shown in FIG.

12 is a flowchart showing the processing steps of a detailed inspection with the configuration shown in FIG.

[Explanation of symbols]

 1 Sand type 2 Sand type defective part 10 Inspection image input part 12 Image processing part 14 Model image storage part 16 Inspection range control part 18 CCD camera (imaging part) 20 Imaging zoom mechanism 22 Projector (light source part) 24 Light source zoom mechanism

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G06T 7/00 G06F 15/62 415 1/00 15/64 M

Claims (4)

[Claims] 1. An inspection image input section for irradiating a three-dimensional object to be inspected with a light cutting line and imaging the surface of the three-dimensional object, an inspection object image output from this inspection image input section, and a predetermined standard model specifying signal. An image processing unit that extracts the non-matching portion with the standard model image specified based on the above and determines the presence or absence of a defect of the three-dimensional object to be inspected based on the non-matching portion data, and the three-dimensional object to be inspected A model image storage unit that stores a plurality of standard model images captured in advance for each of the images, and the inspection image input unit has a light source zoom mechanism for reducing the optical cutting line based on a predetermined inspection range signal. And an imaging unit having an imaging zoom mechanism for enlarging the image of the surface to be inspected based on the predetermined inspection range signal, and the image is input to the inspection image input unit from the outside. An inspection range that outputs a standard model specifying signal to the image processing unit according to an inspection target identification number for inspection target identification and outputs inspection range information in the standard model image to the light source unit and the imaging unit as an inspection range signal. A shape inspection device for a three-dimensional object, which is equipped with a control unit. 2. The inspection range control unit controls the illuminance of the light source unit according to the size of the inspection range when outputting the inspection range signal according to the inspection target identification number to the light source unit. The shape inspection device for a three-dimensional object according to claim 1, further comprising a light amount control function. 3. The image processing unit includes a defect portion calculation function for calculating the generation position and the size of the defect generated in the three-dimensional object to be inspected based on the mismatched portion data, and the inspection range control 3. The section has a detailed inspection control function of outputting an inspection range signal to the light source section and the imaging section based on the defect occurrence position information calculated by the defect part calculation function. 3D object shape inspection device. 4. The inspection image input section is provided with a positioning section that supports the imaging section and the light source section and moves the imaging section and the light source section in the horizontal direction based on a predetermined positioning signal. A divided inspection area specifying function for the control unit to specify in which divided inspection area among the divided inspection areas in which the inspection range is divided in advance based on the defect occurrence position information, and this divided inspection area specifying function The shape inspection apparatus according to claim 1, further comprising a positioning function that outputs a positioning signal to the positioning unit based on the position information of the divided area specified by.