JP2010117182A - Shape measuring apparatus and defect inspecting method - Google Patents

Abstract

A shape measuring apparatus capable of eliminating the influence of external light and shortening the processing time is provided.
A laser projector A, a photoelectric conversion surface of a digital camera B, and an optical lens are arranged in accordance with the Shine-Fluk conditions. The measurement object M is irradiated with slit-shaped laser light from the laser projector A, and the measurement object M is photographed with the digital camera B to acquire image data. This image data is processed with a one-dimensional mask by the mask processing means 23 to remove external light, and the surface shape is extracted from the image data. Thereafter, the three-dimensional shape data generating means 27 generates three-dimensional shape data from a plurality of surface shape data.
[Selection] Figure 5

Description

  The present invention relates to a shape measuring apparatus and a three-dimensional object defect inspection method, and more specifically, irradiates a measurement target with slit-shaped laser light, images the measurement target with an imaging unit, and processes the captured image data It is related with the technique which produces | generates the three-dimensional shape data of a measurement object by performing.

  As similar to the shape measuring apparatus configured as described above, Patent Document 1 relates to a target by irradiating a target with a beam or a slit beam and measuring the position of the reflected light by triangulation with an optical sensor. A signal processing method for obtaining a three-dimensional shape is shown.

  Specifically, a positive signal having a width corresponding to the delay time and a negative signal are obtained by taking a difference signal between the input signal and a delayed signal obtained by delaying the input signal for a predetermined time. The positive signal and the negative signal exist at positions shifted by a delay time corresponding to the reflected light of the pulsed light, and only the signal corresponding to the positive signal exists as a negative signal. Therefore, by obtaining a logical product of the positive signal and the negative signal, only the reflected light of the light beam from the laser light source is separated, and the object can be recognized by removing the influence of noise such as background light. Is.

  In addition, as similar to the shape measuring apparatus, Patent Document 2 discloses that light emitted by measurement illumination is reflected by a measurement object and time until it reaches the TOF sensor is measured by a timer. A distance measuring device that measures the distance from a time and the speed of light to a measurement object and outputs it as an image is shown.

  Specifically, the measurement illumination is pulsed, the received light data of the TOF sensor when the measurement object is not irradiated with light is stored as background light information, and the signal processing circuit subtracts the background information from the measurement light information. The distance difference is calculated as a difference, and the influence of background light is eliminated.

Japanese Patent Publication No. 5-52546 (Example, FIGS. 1 to 5) JP 2008-122223 A (paragraph numbers [0022] to [0037], FIGS. 1 to 11)

  As a process for generating three-dimensional shape data by irradiating a measurement target with laser light, the measurement target irradiated with the laser light is photographed by an imaging means, and laser light is irradiated from the image data photographed by this imaging. It is possible to envisage a process of extracting the surface shape of the measurement object based on the principle of triangulation based on the information obtained by this extraction.

  Since shape measuring devices that measure the shape of parts are generally installed indoors, when acquiring the surface shape by shooting the measurement object, the object to be measured is either indoor lighting or sunlight shining through a window. Often, the brightness of the surface of an object increases. For this reason, it is necessary to remove the influence of external light when extracting a portion irradiated with laser light based on luminance. For this reason, light (external light) other than laser light is irradiated by removing noise as described in Patent Document 1 or by comparing with image data obtained by capturing a background as described in Patent Document 2. The process of excluding the information of the part that has been performed has been performed.

  However, in the one shown in Patent Document 1, there is a face that it is difficult to completely remove the background light, and in the one shown in Patent Document 2, the image data at the timing when the light beam is not irradiated, the timing when the light beam is irradiated, Since the processing for photographing the two types of image data is performed, the time for photographing is disadvantageously increased, and a relatively large storage area is required to store the two types of image data. there were.

  Here, in consideration of the process of irradiating the measurement target with slit-shaped laser light and extracting the portion irradiated with the laser light from the image data obtained by photographing the measurement target, in reality, measurement is performed on the captured image data. It has been confirmed that the luminance of a part of the irradiation area formed on the surface of the object is lowered. That is, in the image data, the laser beam appears as a line having a shape corresponding to the outer surface shape of the measurement object, and a region where the luminance decreases can be confirmed in a part of this line.

  This phenomenon is caused by the fact that when the surface of the measurement object is photographed by the imaging means, focusing corresponding to a wide area of the surface cannot be performed (the focus is not achieved) by the optical lens of the imaging means.

  As described above, when the luminance of the laser light irradiation portion in the image data is reduced due to the optical system, it becomes difficult to extract the portion irradiated with the laser light, and there is room for improvement.

  An object of the present invention is to rationally configure a shape measuring apparatus capable of reliably extracting a shape by eliminating the influence of external light.

  A feature of the present invention is based on a laser projector that irradiates a slit-shaped laser beam, an imaging unit that images a measurement object, and a line irradiation area that is irradiated with laser light from image data captured by the imaging unit. Surface shape extracting means for extracting the surface shape of the measurement object, and an image existing on the optical path surface of the laser beam sent out from the laser projector with a spread is connected to the photoelectric conversion surface of the imaging means. The optical path surface, the photoelectric conversion surface of the image pickup means, and the optical lens of the image pickup means are arranged in accordance with the Shine-Fluk condition so as to image.

  According to this configuration, a line irradiation region is formed along the shape of the outer surface of the measurement target at a portion of the measurement target that is irradiated with the laser light. Since this line irradiation area exists on the optical path surface, and the photoelectric conversion surface of the image pickup means and the optical lens are arranged according to the Shine-Fluk conditions, the image existing on this optical path surface is the photoelectric conversion surface. To form an image. As a result, since all the images of the line irradiation area formed in the image data captured by the imaging means are in focus and in a state of high brightness, the surface shape extraction means can measure the object based on the luminance of the line irradiation area. It is possible to accurately extract the outer shape of the object. As a result, a shape measuring apparatus capable of reliably extracting the shape by eliminating the influence of external light was constructed.

  The present invention increases the luminance of the pixels in which the line irradiation area is formed in the processed image data by processing the image data with a one-dimensional mask before the processing by the surface shape extraction means. A mask processing means for reducing the luminance of pixels other than the line irradiation region may be provided. According to this configuration, the mask processing means performs processing with the one-dimensional mask, thereby increasing the luminance of the line irradiation region irradiated with the laser light in the image data and simultaneously decreasing the luminance of the pixels other than the line irradiation region. It is possible to eliminate the influence of external light.

  In the present invention, the one-dimensional mask is set with mask data having a structure in which a mask value of a target pixel is larger than a mask value of a neighboring pixel, and the mask processing means includes a slit of a slit irradiated with laser light in the image data. A direction orthogonal to the direction is set as a processing target pixel column, and each pixel of the processing target pixel column is set as a target pixel, and the luminance value of the target pixel is converted based on the mask value of the one-dimensional mask. The line with the highest luminance value may be used as the line irradiation region. According to this configuration, it is possible to further increase the luminance of pixels with high luminance and at the same time lower the luminance of pixels other than the line irradiation region, thereby removing noise and clarifying the line irradiation region.

  The present invention includes a photographing control means for photographing a measurement object from different positions by relative movement of the imaging means and the measurement object, and the imaging means obtains a plurality of image data under the control of the photographing control means. The surface shape extraction means extracts the surface shape from a plurality of image data acquired by the imaging means, and generates a three-dimensional data generation means for generating the three-dimensional shape data of the measurement object from the plurality of surface shapes. You may prepare. According to this configuration, the three-dimensional data generation unit can generate three-dimensional shape data of the measurement object based on a plurality of surface shapes.

  According to the present invention, an imaging unit including the laser projector, the optical lens, and the imaging unit is provided at a distal end portion of a robot arm, and the imaging control unit moves the imaging unit along an outer surface of the measurement object. May be. According to this configuration, it is possible to acquire continuous image data while moving the photographing unit along the outer surface of the measurement target.

  According to the present invention, after acquiring the three-dimensional shape data generated by the three-dimensional data generating means of claim 3, the inspection processing means compares with the reference shape data set in advance, and then the defect determination processing means In the above comparison, a part having a dimensional error exceeding a set value may be specified as a defective part.

  According to this configuration, the three-dimensional shape data acquired by the three-dimensional data generating means is compared with the reference shape data, and thereafter, a part having a dimensional error exceeding the set value can be specified as a defective part.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
〔overall structure〕
As shown in FIGS. 1 to 4, an imaging unit Cam is configured in which a laser projector A that irradiates a slit-shaped laser beam and a digital camera B as an imaging unit that images the measurement object M are housed in a case. In addition, an articulated robot arm R that supports the imaging unit Cam at the tip, a robot arm control unit 2 that controls the robot arm R, and image processing of the image data captured by the imaging unit Cam are used to perform triangle processing. The shape measuring device is configured to include an image processing unit 3 that generates three-dimensional shape data of the measurement object M based on the principle of surveying.

  This shape measuring apparatus has a function of displaying the three-dimensional shape data on the monitor 4 and compares the three-dimensional shape data generated by the image processing unit 3 with the three-dimensional shape data of the reference measurement object M. It also functions as a defect inspection device that displays the quality of the shape on the monitor 4 based on the comparison result.

  The shape measurement apparatus moves the imaging unit Cam around the measurement object M by the operation of the robot arm R by executing the process after setting the measurement object M on the measurement table S. Imaging is performed by irradiating the measurement target M with the slit-shaped laser light L from the imaging unit Cam at a plurality of imaging points set in the movement path.

  Image data is acquired by this imaging, and the surface shape of the part irradiated with the laser light L is extracted from the image data and stored, and the measurement object M is based on a plurality of surface shapes stored by continuous imaging. The three-dimensional shape data is generated. Thereafter, the quality of the shape is determined by comparison with the reference three-dimensional shape data of the measurement object.

  In particular, the defect inspection apparatus is designed to take out a product flowing through a line in a press production line, a die-cast production line, etc., measure the three-dimensional shape data of the product, and design the three-dimensional shape data. It is used for inspection to determine the quality of the product by comparing with reference three-dimensional shape data.

  The laser projector A of the imaging unit Cam includes a laser light source 11 and a cylindrical lens 12 that sends out the laser light L from the laser light source 11 in a slit-like shape that expands into a fan shape. In addition, the digital camera B of the imaging unit Cam includes an optical filter 14 that selectively transmits light having a wavelength of the laser light L, an optical lens 15, a focusing mechanism 16 that focuses the optical lens 15, and a CCD or CMOS. Etc., and a photoelectric conversion unit 17 composed of the same.

  The digital camera B only needs to have sensitivity characteristics for efficiently performing photoelectric conversion of the light beam having the wavelength of the laser beam L, and image data of the three primary colors R (red), G (green), and B (blue) are not necessarily required. You don't need anything to do the output. Moreover, in order to send out the laser beam L from the laser light source 11 in a slit shape, a configuration in which the laser beam L can be irradiated onto the fan-shaped region by tilting the mirror or rotating the mirror may be provided.

  In this imaging unit Cam, the optical path surface so that an image existing on the optical path surface LS of the laser light L sent out from the laser projector A with a spread is formed on the photoelectric conversion surface of the photoelectric conversion unit 17 of the digital camera B. The LS, the photoelectric conversion surface, and the optical lens 15 are arranged according to the Shine-Fluk conditions.

  Specifically, a first imaginary line P1 obtained by extending the projection axis Y that sends out the laser beam L from the laser projector A in the counter-projection direction, and a second imaginary line that is orthogonal to the optical axis X of the optical lens 15. Each is arranged according to the Shine-Fluk condition so that P2 and the third virtual straight line P3 obtained by extending the photoelectric conversion surface of the photoelectric conversion unit 17 intersect at one intersection point Q. The first virtual straight line P1, the second virtual straight line P2, and the third virtual straight line P3 exist on a virtual plane in a posture orthogonal to the spreading direction of the laser light L.

  By arranging the laser projector A, the optical lens 15 and the photoelectric conversion unit 17 in accordance with the Shine-Fluk conditions, an image existing on the optical path surface LS of the laser light L is always on the photoelectric conversion surface of the photoelectric conversion unit 17. The image is formed in focus. For this reason, for example, as shown in FIG. 4, when the laser beam L is irradiated from the imaging unit Cam from the oblique information to the measurement object M having a shape with a recess, the recess is formed from the upper surface of the measurement object M. The laser beam L is irradiated to the bottom of the.

  The laser light L irradiated in this way forms a line irradiation region Lm irradiated in a line from the upper surface of the measurement object M to the bottom surface of the recess. As described above, since the image existing on the optical path surface LS of the laser light L is always formed with the photoelectric conversion surface of the photoelectric conversion unit 17 in focus, this line irradiation region Lm has a deep recess. Even if the shape is complicated, the image is formed with the photoelectric conversion surface in focus.

  Further, in the image data G acquired by photographing the measurement object M with the digital camera B, a line irradiation region Lm is formed in the image data G as shown in FIG. 4, and the pixels corresponding to the line irradiation region Lm are The brightness is high.

  The robot arm R has a structure in which a plurality of arm portions Ra are connected at joints, and each arm portion Ra can freely swing at the joints. A cable 5 for controlling the imaging unit Cam provided in the robot arm R and acquiring image data is provided between the imaging unit Cam and the image processing unit 3. A cable 6 is provided between the robot arm control unit 2 and the image processing unit 3 for outputting a control signal to the robot arm control unit 2 and transmitting a signal indicating that the imaging unit Cam has reached the shooting point. .

(Image processing unit)
The image processing unit 3 is composed of a general-purpose computer, and this general-purpose computer includes a processing system shown in FIG. The processing system is assumed to be configured by software, but may be configured by hardware, or may be configured by a combination of software and hardware.

  That is, the image processing unit 3 includes the reference shape data acquisition unit 20, the imaging control unit 21, the image data acquisition unit 22, the mask processing unit 23, the surface shape extraction unit 24, the distortion correction processing unit 25, the surface A shape storage unit 26, a three-dimensional shape data generation unit 27, an inspection processing unit 28, and a defect determination processing unit 29 are provided.

  The reference shape data acquisition means 20 performs processing for acquiring reference shape data which is ideal three-dimensional shape data of the measurement object M, and this acquisition is acquired via a medium or via a communication line. Is assumed to be processed.

  The imaging control unit 21 sets a plurality of imaging points based on the reference shape data acquired by the reference shape data acquisition unit 20 and outputs coordinate data indicating the imaging points to the robot arm control unit 2. In addition, the shooting control unit 21 performs shooting by controlling the digital camera B of the imaging unit Cam in a state where the imaging unit Cam is set as a shooting point.

  The image data acquisition unit 22 acquires image data from the photoelectric conversion unit 17 of the digital camera B when the imaging unit Cam reaches the shooting point and shooting is performed.

  The mask processing unit 23 performs a process of specifying a region (line irradiation region Lm) irradiated with the laser light L in the image data by masking the image data. In this process, the one-dimensional mask 23m shown in FIG. 6 is used, and as shown in FIG. 7A, the direction orthogonal to the direction of the slit irradiated with the laser light L is set in the processing target pixel row RD. The individual pixel of the processing target pixel row RD is set as the target pixel Px, the luminance value of the target pixel Px is converted based on the mask value of the one-dimensional mask 23m, and the position of the pixel having the highest converted luminance value is the line irradiation region. Processing to set Lm is performed.

  The mask value of the one-dimensional mask 23m is set as indicated by a numerical value in FIG. 6, and has a data structure that is smaller as the value of the separated position is based on the mask value corresponding to the target pixel Px. Using this one-dimensional mask 23m, the product of the luminance value (density value) of the target pixel Px and the luminance value (density value) of the pixel located at the position connected to the target pixel Px is obtained, and the integration of these product values is further performed. A product-sum process for replacing the value with the luminance value of the target pixel Px is performed for each successive pixel.

  Taking the image data G of one frame (one frame) shown in FIG. 7A as an example, the luminance value of each pixel in the processing target pixel row RD can be shown as shown in FIG. 7B. In this image data, the luminance value of each pixel is converted by the one-dimensional mask 23m, and the processing by the one-dimensional mask 23m is performed on the pixel row adjacent to the processing target pixel row RD (in FIG. 7A, on the upper side or the lower side). ) In order. As a result, processing for all the pixels of one frame of image data is performed.

  As a result of this processing, as shown in FIGS. 8A and 8B, the luminance of the line irradiation region Lm is increased, and external light that has been irradiated on the background can be eliminated.

  The surface shape extraction means 24 identifies the line irradiation region Lm based on the luminance in the image data that has been processed by the mask processing means 23, and extracts the surface shape based on the triangulation principle from the identified position. In this process, the line irradiation region Lm irradiated with the laser light L is displaced in the image data in accordance with the relative distance between the digital camera B and the measurement object M, and therefore the laser light in the image data. While specifying the irradiation position of L, the surface shape is acquired by acquiring the position of the imaging unit Cam from the robot arm control unit 2.

  In the distortion correction processing means 25, a process for correcting distortion caused by the photoelectric conversion surface of the photoelectric conversion unit 17 and the optical lens 15 being arranged according to the Shine-Fluk conditions is performed. In other words, correction is made to correct the deformation caused by photographing under the condition that the optical axis X of the optical lens 15 is not orthogonal to the surface of the measurement object M.

  The surface shape storage unit 26 accesses data in a storage such as a semiconductor memory. When storing the surface shape in the storage, the surface shape storage unit 26 acquires the position of the imaging unit Cam from the robot arm control unit 2. The surface shape data and the photographing position data are stored in association with each other.

  The three-dimensional shape data generation unit 27 generates three-dimensional shape data corresponding to the shape of the measurement object M from the plurality of surface shape data stored in the surface shape storage unit 26. At the time of generation, processing in a form in which surface shape data is continuously arranged in a three-dimensional space is performed based on imaging position data stored in association with the surface shape data. The three-dimensional shape data generated in this way is displayed on the monitor 4.

  The inspection processing unit 28 compares the reference shape data acquired by the reference shape data acquisition unit 20 with the three-dimensional shape data generated by the three-dimensional shape data generation unit 27, so that parts having different outer surface shapes are obtained. Inspect for presence / absence.

  The defect determination processing means 29 identifies a part having a dimensional error that exceeds the set value as a defective part by the inspection process in the inspection processing part 28, makes the hue of the defective part different, blinks, etc. Is displayed on the monitor 4.

[Process type]
An outline of processing in the image processing unit 3 is shown in a flowchart of FIG. That is, with the measurement object M set on the measurement table S, the reference shape data acquisition unit 20 of the image processing unit 3 acquires the reference shape data of the measurement object M and stores it in a storage such as a memory. (Step # 01).

  Next, the robot arm R is controlled to move the image pickup unit Cam to the shooting point, and shooting is performed to acquire one frame of image data, mask processing, acquisition of surface shape data, distortion correction, surface shape data. Is stored (steps # 02 to # 06).

  The shooting point is set by the shooting control unit 21 based on the reference shape data, and the shooting control unit 21 moves the imaging unit Cam along the outer surface of the measurement object M with respect to the robot arm control unit 2. Give control data.

  When the imaging unit Cam reaches the shooting point, the shooting control unit 21 controls the digital camera B to perform shooting, and the image data acquisition unit 22 acquires one frame of image data.

  In addition, the mask processing unit 23 sets the processing target pixel row RD for the acquired image data of one frame, and performs processing based on the one-dimensional mask 23m, thereby eliminating external light of the image data. Further, the surface shape extraction unit 24 extracts the surface shape from the image data from which the external light is excluded, the distortion correction processing unit 25 corrects the distortion, and the surface shape storage unit 26 stores the surface shape data on the storage. The shooting position data is stored in association with each other.

  This series of processing is performed before the next image data is acquired (captured), and after this series of processing, it is determined that shooting at all shooting points has not been completed. In this case, the imaging unit Cam is moved to the next shooting point, and the processes of # 02 to # 06 described above are performed. If it is determined that the process is completed, the process proceeds to the next step (# 07 step). ).

  When it is determined that the photographing is completed, the robot arm R is moved to the home position, and the three-dimensional shape data generation unit 27 generates the three-dimensional shape data based on the data of the plurality of surface shapes, The generated three-dimensional image data is displayed on the monitor 4 (Step # 09).

  When generating the three-dimensional shape data, a plurality of surface shape data are continuously arranged in the three-dimensional space on the memory based on the photographing position data stored in association with the three-dimensional shape data. Form processing is performed.

  After the three-dimensional shape data is generated in this manner, the inspection processing unit 28 inspects the reference shape data and the three-dimensional shape data by comparison, and the defect determination processing unit determines whether there is a defect (# 10). , # 11 step).

  The inspection processing unit 28 determines whether or not there is a part having a different shape. When it is determined in the inspection processing means 28 that there is a part having a different shape, if the difference in dimensions in the part having a different shape exceeds a set value, the shape is different in order to improve the distinguishability of the part. The display is displayed on the monitor 4 in a display form such as making the hues of the parts different or blinking.

[Example effects]
As described above, according to the present invention, the laser projector A, the photoelectric conversion surface of the photoelectric conversion unit 17 of the digital camera B, and the optical lens 15 are arranged in accordance with the Shine-Flux conditions. The image existing in the LS is always formed in a state where the photoelectric conversion surface of the photoelectric conversion unit 17 is in focus. Accordingly, in the captured image data, the line irradiation region Lm by the laser light L irradiated to the measurement object M is clear and has high brightness. Furthermore, the mask processing means 23 performs processing with the one-dimensional mask 23m, so that the influence of external light can be eliminated and the brightness of the portion irradiated with the laser light L from the image data can be further increased.

  As a result, the effect of outside light such as indoor lighting or sunlight shining from a window is compared with that for simply extracting a portion irradiated with laser light L from image data based on the luminance of the pixel. It is possible to accurately extract the line irradiation region Lm irradiated with the laser light L and obtain the surface shape data by the principle of triangulation. Further, the surface shape data obtained in this way is taken by an optical system that complies with the Shine-Fluk conditions, and therefore includes distortion. For this reason, it can be acquired as surface shape data whose distortion has been corrected by the correction processing in the distortion correction processing means 25.

  Further, the corrected surface shape data is stored in a storage such as a memory in association with the photographing position data. In addition, since processing with a large weight such as delay processing and comparison processing for obtaining the difference between a plurality of image data is not performed in the processing until acquiring surface shape data, processing from photographing to storage of surface shape data is not performed. Can be completed before the next frame is shot, continuous shooting of the image data is realized, and the time until the shooting of the measurement object M is completed can be shortened.

  Next, when the three-dimensional shape data generation unit 27 generates the three-dimensional shape data of the measurement object M from the plurality of surface shape data, the photographing position stored in association with each of the plurality of surface shape data. Based on the data, the surface shape data is arranged in a three-dimensional space.

  The inspection processing unit 28 compares the reference data of the measurement object M with the three-dimensional shape data generated by the three-dimensional shape data generation unit 27, and as a result of the comparison, the dimension error of a part having a different shape exceeds the set value. Since the defect determination processing means 29 displays on the monitor 4 in a state where the identification of the defective part is enhanced, the defective part can be easily grasped.

The perspective view which shows the whole shape measuring device Schematic diagram showing the arrangement of the laser projector, optical lens, and photoelectric conversion unit The perspective view which shows the structure of an imaging unit The perspective view which shows the imaging | photography form of the line irradiation area | region in which a measurement target object is formed Block diagram of control system The figure which shows the processing form by a mask processing means The figure which shows the image data and luminance value before mask processing The figure which shows the image data and luminance value after a mask process Flow chart showing processing mode

Explanation of symbols

DESCRIPTION OF SYMBOLS 15 Optical lens 21 Imaging | photography control means 23 Mask processing means 23m One-dimensional mask 24 Surface shape extraction means 27 Three-dimensional data generation means 28 Inspection processing means 29 Defect determination processing means A Laser projector B Imaging means (digital camera)
Cam Imaging unit L Laser light LF Optical path surface Lm Line irradiation area M Measurement object R Robot arm RD Processing target pixel row

Claims (6)

A laser projector that irradiates a slit-shaped laser beam, an imaging unit that images the measurement object, and a line irradiation area irradiated with laser light from the image data captured by the imaging unit. A surface shape extracting means for extracting the surface shape,
The optical path surface, the photoelectric conversion surface of the imaging means, and the imaging so that an image existing on the optical path surface of the laser light sent out from the laser projector with spread is formed on the photoelectric conversion surface of the imaging means A shape measuring device in which the optical lens of the means is arranged according to the conditions of Shine Fluke.   Before the processing by the surface shape extraction means, the image data is processed with a one-dimensional mask to increase the luminance of the pixels in which the line irradiation area is formed in the processed image data, and the line irradiation The shape measuring apparatus according to claim 1, further comprising a mask processing unit that weakens the luminance of pixels other than the region. The one-dimensional mask is set with mask data having a structure in which the mask value of the target pixel is larger than the mask values of neighboring pixels,
The mask processing means sets a direction orthogonal to the direction of the slit irradiated with laser light in the image data as a processing target pixel column, and sets each pixel of the processing target pixel column as a target pixel. The shape measuring apparatus according to claim 2, wherein a luminance value is converted based on a mask value of a one-dimensional mask, and the line irradiation region has a highest luminance value after conversion. The image pickup means and the measurement object are provided with shooting control means for shooting the measurement object from different positions by relative movement, and the image pickup means takes a plurality of image data under the control of the shooting control means. The surface shape extraction means extracts the surface shape from a plurality of image data acquired by the imaging means,
The shape measuring apparatus according to any one of claims 1 to 3, further comprising a three-dimensional data generation unit that generates three-dimensional shape data of a measurement object from a plurality of surface shapes.   The imaging unit comprising the laser projector, the optical lens, and the imaging unit is provided at a tip of a robot arm, and the imaging control unit moves the imaging unit along an outer surface of the measurement object. Shape measuring device.   5. After obtaining the three-dimensional shape data generated by the three-dimensional data generation means according to claim 4, the inspection processing means compares with the reference shape data set in advance, and then the defect determination processing means performs the comparison. A defect inspection method for specifying a defect portion where a dimensional error of a portion having a different shape exceeds a set value.