JP5728699B2 - Surface inspection apparatus, surface inspection method, and surface inspection program - Google Patents

Description

  The present invention provides non-contact inspection of surface defects such as rubbing, stepping stones, paint peeling, fading and dents on objects subject to strong surface reflections, such as by applying glossy paint to automobile bodies and parts. The present invention relates to a surface inspection apparatus, a surface inspection method, and a surface inspection program.

  Frontal inspection of an object using an imaging device that captures an image of the target object and image processing technology has been widely used in the industry in recent years because it can grasp the state of the target object with high speed and high accuracy without contact. Yes. However, since the body of an automobile has a smooth surface and a gloss coating, the specular reflection is severe, which is disadvantageous for image measurement. In addition, the shape of the body of an automobile varies depending on the part, and the colors of painting are also various. Therefore, there is a demand for a surface inspection apparatus that can cope with various shapes, surface colors, and reflection characteristics.

  In a conventionally proposed apparatus for inspecting scratches or paint on the body of an automobile, a technique of irradiating a measurement site with a bright and dark stripe pattern and capturing a reflection pattern image is the mainstream. For example, Patent Document 1 includes a light / dark pattern irradiating unit that irradiates a predetermined light / dark pattern on an outer peripheral surface of a measurement object, and an imaging unit that images the outer peripheral surface of the measurement object. A coating defect inspection apparatus is disclosed in which a time-series image is obtained by irradiating and capturing a light and dark pattern while moving a vehicle on a line for measurement, and a coating defect is detected from the time-series image.

  Patent Document 2 discloses a coating defect inspection apparatus using a stripe-type light and dark pattern as in Patent Document 1, and includes a light irradiation unit and an imaging unit using two arm robots without moving an automobile. A coating defect inspection device that captures an image of the entire measurement object by moving it is disclosed.

  In the case of an apparatus that irradiates and inspects a stripe pattern, such as the coating defect inspection apparatus described in Patent Documents 1 and 2, it is necessary to irradiate the entire measurement object with a stripe-type light and dark pattern. Similarly, it is necessary to image the entire measurement object in order to image the illuminated light / dark pattern. However, since there are restrictions on the irradiation range of the light irradiation means and the imaging range of the imaging means, the illumination and imaging of the light / dark pattern can be performed only in a part of the local area at a time.

  Therefore, when a coating defect is inspected over the entire measurement object using such an inspection apparatus, a partial light pattern is irradiated and imaged over the entire measurement object, and a time-series image is obtained. It is necessary to acquire and perform complex image processing based on this time-series image. For this reason, the inspection apparatus is complicated, the measurement time is long, and the cost is high.

  On the other hand, in the case of a defect accompanying a shape change such as a dent, the surface color characteristics may or may not change when a bright and dark stripe pattern is irradiated. Recesses that do not involve changes in surface color characteristics cannot be detected by a two-dimensional inspection method based on general color information analysis, and it is necessary to use a three-dimensional measurement technique that can detect shape changes.

  In general, three-dimensional image measurement is performed by irradiating the measurement target with a passive type that performs measurement without irradiating the measurement target with specific light or radio waves that assist measurement, and with light, sound waves, or radio waves. The active type that performs measurement using the information is classified.

  A typical passive method uses two cameras to capture images of a measurement object from different viewpoints, such as human eyes, and calculates the three-dimensional shape of the measurement object based on the principle of triangulation. Stereo viewing method. Stereoscopic three-dimensional shape measurement method is a method that can measure the three-dimensional shape of an object to be measured using only two or more cameras, and measures the three-dimensional shape of an object using simple equipment. There are advantages you can do.

  However, in such a passive method, so-called association is required in which measurement points and corresponding points are extracted from a plurality of images. For this reason, a dent without the color change has no characteristics and cannot be measured. It is also difficult to measure complicated dents in correspondence with color changes.

  On the other hand, as an active three-dimensional information measurement method, for example, a method by pattern light projection described in Patent Document 3 is known. In the three-dimensional information measurement method using pattern light projection, pattern light is projected onto a measurement object, and the reflection pattern is analyzed to obtain three-dimensional information on the surface shape of the object. There is a feature that measurement is also possible.

FIG. 6 is a geometric relationship diagram showing the basic principle of a conventional three-dimensional information measurement method using pattern light projection. In FIG. 6, O _P is the lens center of the pattern light projector, O _C is the lens center of the observation camera. The pattern light projector and the camera are spaced apart by a certain distance b. The pattern light projector projects the measurement pattern light onto the measurement point M with an angle of β. On the other hand, the observation camera observes the measurement point M and captures an image. The observation angle at this time is α.

The depth coordinate Z in the three-dimensional space world coordinate system (O, X, Y, Z) of the measurement point M at this time is obtained as follows.

  α is the observation angle (the viewing angle of the measurement point as seen from the observation camera), β is the projection angle (the direction angle of the light projected from the pattern light projector onto the measurement object), and b is the pattern from the center of the observation camera This is the distance to the center of the optical projector.

Japanese Patent No. 3211681 Japanese Patent No. 4014027 JP 2006-145405 A

  However, as described above, when the object to be measured is an object having a strong surface reflection such as the body of an automobile, when pattern light is projected onto the object to be measured, highlights are formed in the part that is regularly reflected by the observation camera. Pattern light is reflected very strongly. On the other hand, the intensity of the reflected light is extremely weak compared to the highlight in the portion other than the regular reflection, and it is difficult to image the reflection pattern necessary for the three-dimensional information measurement. For this reason, in the case of an object with strong surface reflection, there is a problem in that three-dimensional measurement cannot be performed because the reflection pattern of the measurement object cannot be imaged.

  Therefore, in the present invention, even a measurement object having a strong surface reflection due to the glossy coating of the body of the automobile or its parts, etc., the surface of the surface can be accurately and quickly at a low cost with a simple structure. An object of the present invention is to provide a surface inspection apparatus, a surface inspection program, and a surface inspection method capable of inspecting surface defects such as wrinkles and paint peeling and dents.

  A surface inspection apparatus according to the present invention includes a measurement environment configuration apparatus that configures a measurement space having a curved wall, and an illumination apparatus that uniformly illuminates and illuminates a measurement target in the measurement space with light source light. In order to project the pattern light on the curved wall so that it does not directly hit the measurement object and reflect it, and to project the reflected pattern light (hereinafter referred to as “reflection pattern light”) onto the measurement object Pattern light projection device, imaging device that captures all illumination image of measurement object illuminated by illumination device, and reflection pattern image of measurement object onto which reflection pattern light is projected, all illumination image and reflection pattern And a data processing device for detecting a surface defect of the measurement object from the image.

  In addition, the surface inspection method of the present invention includes forming a measurement space having a curved wall, illuminating the measurement object in the measurement space with the light source light uniformly dimmed, and the illuminated measurement object. Take a picture of the entire illumination of the object, project the pattern light onto the curved wall so that it does not directly hit the measurement object, reflect it, and project the reflected pattern light (reflection pattern light) onto the measurement object Capturing a reflection pattern image of the measurement object onto which the reflection pattern light is projected, and detecting a surface defect of the measurement object from the entire illumination image and the reflection pattern image.

  According to these inventions, a measurement space having a curved wall is formed, and the measurement object in the measurement space is illuminated with the light source light uniformly dimmed, and the illuminated measurement object is fully illuminated. The pattern light is projected and reflected on the curved wall so that it does not directly strike the measurement object, and the reflected pattern light (reflection pattern light) is projected onto the measurement object to reflect the pattern light. Since the reflection pattern image of the measurement object on which is projected is captured, the surface reflection is difficult to be captured. Therefore, even if the object to be measured has a strong surface reflection due to the glossy coating of the body of the car or its parts, the reflected pattern light of the object to be measured can be imaged, and the entire illumination image And the surface defect of the measurement object can be detected from the reflection pattern image.

  Here, the curved wall is formed of a semi-transparent material, and the light source light is provided outside the curved wall, and passes through the curved wall and is illuminated on the measurement object. It is desirable to be a thing. As a result, the light source light is uniformly dimmed when passing through the curved wall formed of a semi-transparent material from the outside of the curved wall, and is illuminated on the measurement object and projected onto the curved wall. The pattern light is uniformly reflected and projected onto the measurement object, and the entire illumination image and the reflection pattern image are captured.

  Moreover, it is desirable that the imaging device captures the reflection pattern image from a viewing angle different from the projection angle of the projection pattern light. By capturing the reflection pattern image from a viewing angle different from the projection angle of the projection pattern light, it is further difficult to capture the surface reflection, and the surface defect of the measurement object can be detected with higher accuracy.

  Further, the surface inspection program of the present invention illuminates the measurement object in the measurement space having the curved wall with the light source light being uniformly reduced, and captures all illumination images of the illuminated measurement object. The pattern light is projected and reflected on the curved wall so that it does not directly strike the measurement object, and the reflected pattern light (reflection pattern light) is projected onto the measurement object. This is for causing a computer to execute imaging of a reflection pattern image of a projected measurement object, and to detect a surface defect of the measurement object from all illumination images and reflection pattern images. According to the computer which executed this program, the same operation and effect as the above-mentioned surface inspection method of the present invention can be produced.

  By capturing all illumination images and reflection pattern images of the measurement object in the measurement space, and detecting the surface defect of the measurement object from the captured image, this image can be obtained simply by shooting the measurement object normally. The surface defect of the measurement object can be easily detected from the captured image, and the surface defect such as surface scratches, paint peeling, fading and dents can be inspected with high accuracy and high speed at a low cost with a simple structure. Is possible.

1 is an overall configuration diagram showing an automobile surface defect inspection system in an embodiment of the present invention. It is a figure which shows the example of measurement space. It is a block diagram which shows the detailed structure of the data processor of FIG. It is a geometric relationship figure which shows the basic principle of the three-dimensional image analysis in this embodiment. It is a flowchart which shows the flow of the test | inspection by the motor vehicle surface defect inspection system of FIG. It is a geometric relationship figure which shows the basic principle of the three-dimensional information measurement method by the conventional pattern light projection.

  1 is an overall configuration diagram showing an automobile surface defect inspection system according to an embodiment of the present invention, FIG. 2 is a diagram showing an example of a measurement space, and FIG. 3 is a block diagram showing a detailed configuration of the data processing apparatus of FIG. is there.

  In FIG. 1, an automobile surface defect inspection system according to an embodiment of the present invention is a surface inspection apparatus that inspects a surface defect of a measurement object A such as an automobile body or its parts in a non-contact manner. A tent 1a as a measurement environment constituting device that constitutes a measurement space 1 for measurement, illumination devices 2-1, 2-2, 2-3, and the like, a pattern light projection device 3 that projects pattern light, and an imaging device And a data processing device 5 that processes data of an image captured by the camera device 4.

  The illumination devices 2-1, 2-2, 2-3, the pattern light projection device 3 and the camera device 4, and the data processing device 5 are transmission cables 6-1, 6-2 that can transmit each data. 6-3, 6-4, and 6-5, respectively.

  The measurement space 1 has a curved wall 1b. As shown in FIGS. 2A and 2B, the measurement space 1 is formed in a shape having a curved surface, such as a dome shape or a tunnel shape, and the curved surface portion is a curved wall 1b. Is configured. The curved wall 1b is made of a light translucent material such as white, and has predetermined light reflection characteristics and transmission characteristics. In the present embodiment, the curved wall 1b has a reflectance of 20% to 95% and a transmittance of 5% to 80%.

  The illumination devices 2-1, 2-2, and 2-3 are illumination devices such as ordinary incandescent bulbs, fluorescent lamps, halogens, and LEDs (light emitting diodes). The light source light of the illuminating devices 2-1, 2-2 and 2-3 is uniformly dimmed and diffused by the curved wall 1 b and is illuminated on the measurement object A in the measurement space 1.

  The pattern light projection device 3 converts the pattern data necessary for measurement formed by the data processing device 5 into pattern light so that the measurement object A does not exist, that is, does not directly hit the measurement object A. It is a device that projects onto the curved wall 1b. The pattern light projected on the curved wall 1b is reflected on the curved wall 1b, and the reflected pattern light (hereinafter referred to as “reflection pattern light”) is projected on the measurement object A. As the pattern light projector 3, a commercially available simple device such as a liquid crystal projector, a DLP (Digital Light Processing (trademark)) projector, a laser light projector, an LED (semiconductor) projector, or a film projector can be used.

  The camera device 4 is a digital camera. The camera device 4 may be any digital camera such as an 8-bit, 10-bit, 12-bit, or 16-bit camera, a CCD, a CMOS, a still image camera, a video camera, a video camera, or the like. . The camera device 4 is arranged so that the optical axis (viewing angle) of the lens is not the same as the projection direction (projection angle) of the pattern light (reflection pattern light) reflected by the curved wall 1b. In FIG. 1, only one camera device 4 is illustrated, but a plurality of camera devices 4 are arranged to have different viewing angles.

  The data processing device 5 is a computer that executes a surface inspection program. The data processing apparatus 5 executes the surface inspection program to thereby store the storage means 10, the illumination means 11, the all illumination image imaging means 12, the projection pattern generation means 13, the pattern light projection means 14, and the reflection pattern image imaging means shown in FIG. 15, 3D image synthesis means 16, 2D image analysis means 17, 3D image analysis means 18, defect detection classification size estimation means 19, measurement result expression means 20, and storage output means 21.

  The storage unit 10 includes a projection pattern generated by the projection pattern generation unit 13, data of an image transmitted from the camera device 4, a three-dimensional image synthesis unit 16, a two-dimensional image analysis unit 17, a three-dimensional image analysis unit 18, Data of the results calculated by the respective means such as the defect detection classification size estimating means 19, the measurement result expressing means 20 and the storage output means 21 are stored.

  The illumination unit 11 controls the illumination devices 2-1, 2-2 and 2-3 in order to capture all illumination images by the camera device 4. The light source light of the illuminating devices 2-1, 2-2, 2-3 is uniformly attenuated by the curved wall 1 b of the tent 1 a and is illuminated on the measurement object A in the measurement space 1. The light source light of the lighting devices 2-1, 2-2, and 2-3 may be any color, but is uniform light without a pattern.

  The all-illuminated image capturing unit 12 controls the camera device 4 when the measurement object A is illuminated by the illumination unit 11, thereby illuminating with these illumination devices 2-1, 2-2 and 2-3. The entire illumination image of the measured object A is captured and stored in the storage means 10.

  The projection pattern generation unit 13 generates a measurement projection pattern and stores it in the storage unit 10. As the shape of the measurement projection pattern, a striped or circular pattern can be used, and other shapes are also possible. The color of the pattern can be uniform or different, and its intensity distribution may be uniform or non-uniform. The projection pattern generation means 13 generates a measurement projection pattern having a shape, color and spatial frequency distribution optimal for measurement according to the size, shape and surface color distribution of the measurement object A.

  The pattern light projection unit 14 projects the light of the measurement projection pattern stored in the storage unit 10 by the pattern light projection device 3. The pattern light projected by the pattern light projector 3 is projected onto the curved surface portion of the curved wall 1b of the measurement space 1 so as not to directly hit the measurement object A, and is reflected by the curved surface portion and reflected. Pattern light (hereinafter referred to as “reflection pattern light”) is projected onto the measurement object A.

  The reflection pattern image imaging means 15 captures the reflection pattern image of the measurement object A onto which the reflection pattern light is projected by controlling the camera device 4 when the reflection pattern light is projected onto the measurement object A. And stored in the storage means 10.

The three-dimensional image synthesizing means 16 three-dimensionally transforms all illumination images and reflection pattern images captured from the respective viewing angles by the plurality of camera devices 4 and arranges them in the same posture, and synthesizes them for a wide range of inspections. The image for use is obtained. The obtained inspection image is stored in the storage means 10. Three-dimensional images are synthesized using the three-dimensional spatial coordinate transformation technology shown in Equation (2), and the coordinates of each image are geometrically transformed into a unified three-dimensional coordinate space by three-dimensional rotation and translation. Then synthesize.
X ′ = RX + T (2)
Here, X is a coordinate vector before conversion, X ′ is a coordinate vector after conversion, R is a rotation matrix, and T is a translation matrix.

  The two-dimensional image analysis means 17 analyzes the color intensity of the entire illumination image and the reflection pattern image, and detects an area that is large in the color intensity that is suspected of being defective and that is not a print pattern or a character as the defect candidate area 1. To do. The detected defect candidate area 1 is stored in the storage means 10.

  The three-dimensional image analyzing unit 18 analyzes the combined image of the entire illumination image and the reflection pattern image synthesized by the three-dimensional image synthesizing unit 16 and has a large change in the depth coordinate value Z that is suspected of being a defect and is unique to the body. An area that is not a shape change is detected as a defect candidate area 2. The detected defect candidate area 2 is stored in the storage means 10. The depth coordinate value Z is calculated as follows.

FIG. 4 is a geometric relationship diagram showing the basic principle of the three-dimensional image analysis in the present embodiment.
The curved wall 1b of the measurement space 1 that reflects the projection pattern light is referred to as a straight line connecting the lens center O _P of the pattern light projection device 3 and the lens center O _C of the camera device 4 (hereinafter referred to as “lens center line”). ) Is formed at an angle of φ. φ varies depending on the shape of the measurement space 1 and the position of the reflection point. The pattern light projector 3 projects the pattern light for measurement onto the curved wall 1b with an angle β, and the light reflected from the curved wall 1b is projected onto the measurement object A. The camera device 4 captures an image of the measurement object A, and a measurement point M on the surface of the measurement object A is observed from this image. The observation angle at this time is α.

ここで、αは式（１）と同じく観測角度（カメラから見た計測点の所在の視角）、θは投影角度（パターン光投影装置３から曲面状壁１ｂに投影した光の方向角）、φは曲面状壁１ｂとレンズ中心線との間の角度、ｂはカメラ装置４のレンズ中心Ｏ_Cからパターン光投影装置３のレンズ中心Ｏ_Pまでの距離、ｈはレンズ中心線と曲面状壁１ｂとの間の距離である。 The depth coordinate value Z in the three-dimensional space world coordinate system (O, X, Y, Z) of the measurement point M is obtained as the following equation (3).

Here, α is the observation angle (the viewing angle of the location of the measurement point as seen from the camera) as in the equation (1), θ is the projection angle (the direction angle of the light projected from the pattern light projector 3 onto the curved wall 1b), the angle between the curved wall 1b and the lens centerline phi, b is the distance from the lens center O _C of the camera apparatus 4 to the lens center O _P of the pattern light projection device 3, h is the lens center line curved wall It is the distance between 1b.

  In the defect candidate area 1 and the defect candidate area 2, the defect detection classification size estimation means 19 specifies and classifies defects according to specific characteristics such as pointed wrinkles, line wrinkles, area wrinkles, fading, paint peeling and dents. Is to do. In addition, dimensions such as the length of the line defect, the diameter of the dot defect, the area of the area defect, and the depth of the dent defect are estimated.

  The measurement result expression means 20 expresses the image data captured by the camera device 4 and the defect data detected by the defect detection classification size estimation means 19 by a method such as CAD drawing or computer graphics (CG). .

  The storage output unit 21 stores the image data captured by the camera device 4 and the defect data detected by the defect detection classification size estimation unit 19 as an image file or a text file on a recording medium or prints on a data sheet. It is something to do.

  Next, a surface inspection method using the automobile surface defect inspection system having the above configuration will be described. FIG. 5 is a flowchart showing a flow of inspection by the automobile surface defect inspection system of FIG.

  First, before measurement, the tent 1a, the lighting devices 2-1, 2-2, 2-3, the pattern light projection device 3, the camera device 4, the data processing device 5, and the like are arranged to construct a measurement environment (step S101). ).

  Next, the measurement object A is placed in the measurement space 1 and surface inspection is performed. For measurement, the lighting devices 2-1, 2-2 and 2-3 are first turned on and the measurement object A is irradiated with illumination light (all irradiation) (step S 102). At this time, the illumination light from the illumination devices 2-1, 2-2, 2-3 is not directly irradiated onto the measurement object A, but the illumination light attenuated by the curved wall 1 b of the measurement space 1 is irradiated. Is done.

  Next, all illumination images are captured (step S103). The all-illuminated image capturing unit 12 captures the image of the measurement object A illuminated by the illuminating unit 11, that is, the all-illuminated image, by the image capturing device 4 and transmits it to the data processing device 5. All illumination image data transmitted from the camera device 4 is stored in the storage means 10 by the data processing device 5 and processed.

  The data processing device 5 analyzes the color intensity distribution of all the captured illumination images, and determines whether or not it has an ideal intensity distribution necessary for the object surface defect inspection. If the image is too bright or too dark and the color intensity distribution is not ideal, the parameters of the camera device 4 are corrected and the entire illumination image is captured again (step S104). If the color intensity distribution is ideal, proceed to the next step.

  Next, the pattern light projection unit 14 projects the measurement projection pattern generated by the projection pattern generation unit 13 onto the curved wall 1b of the measurement space 1 by the pattern light projection device 3 (step S105). This projection pattern is reflected by the curved wall 1b and indirectly projected onto the measurement object A (step S106).

  Next, the reflection pattern image imaging means 15 is reflected by the curved surface portion of the curved wall 1b of the measurement space 1 by the pattern light projection device 3, and is reflected on the measurement object A, that is, a reflection pattern image. Is captured by the camera device 4 and transmitted to the data processing device 5 (step S107). The reflection pattern image data transmitted from the camera device 4 is stored in the storage means 10 by the data processing device 5 and processed.

  The data processing device 5 analyzes the color intensity distribution of the captured reflection pattern image and determines whether or not it has an ideal intensity distribution necessary for the object surface defect inspection. If the image is too bright or too dark and the color intensity distribution is not ideal, the parameters of the camera device 4 are corrected and a reflection pattern image is captured again (step S108). If the color intensity distribution is ideal, proceed to the next step.

  Next, the three-dimensional image synthesizing unit 16 three-dimensionally synthesizes all illumination images and reflection pattern images captured from the respective viewing angles by the plurality of camera devices 4 to create a wide range of inspection images (step S109). ).

  Next, the two-dimensional image analysis means 17 analyzes the color intensity of the entire illumination image and the reflection pattern image, and determines an area that has a large change in color intensity that is suspected of being a defect and that is not a print pattern or a character as a defect candidate area. 1 is detected (step S110).

  Next, the three-dimensional image analysis means 18 analyzes the combined image of the entire illumination image and the reflection pattern image, and determines an area where the change of the depth coordinate value Z that is suspected of being a defect is large and is not a shape change inherent to the body. It is detected as a defect candidate area 2 (step S111).

  Next, the defect detection classification size estimation means 19 identifies and classifies defects in the defect candidate area 1 and the defect area candidate 2 according to specific characteristics such as wrinkles, fading, paint peeling and dents. In addition, dimensions (sizes) such as the length of the line defect, the diameter of the dot defect, the area of the area defect, and the depth of the dent defect are estimated (step S112).

  Next, the measurement result expression means 20 displays the detected defect on the monitor of the data processing device 5, puts a different color mark depending on the type of defect, and displays the estimated defect size. Further, according to the user's needs, the measurement object A is divided into several areas, and the number and types of defects in each area, the distribution of defects in the entire measurement object A, and the like are displayed (step S113).

  Finally, the storage output unit 21 stores the image data captured by the camera device 4 and the defect data detected by the defect detection classification size estimation unit 19 as an image file or a text file on a recording medium, or a data sheet. Printing on the top (step S114).

  As described above, according to the automobile surface defect inspection system in the present embodiment, the surface defect of the measurement object A can be easily obtained from the captured image only by photographing the measurement object A in the measurement space 1 normally. With a simple structure, it is possible to inspect surface defects such as surface scratches, paint peeling, fading and dents at a low cost with high accuracy and high speed.

  Moreover, since the camera apparatus 4 images the measurement object A in a state of being relatively stationary with respect to the measurement object A, the measurement object A or the camera apparatus 4 is moved to scan the measurement object A. The measurement object A can be imaged in a short time. Therefore, an apparatus necessary for scanning is unnecessary, and it is possible to shorten the inspection time, improve the inspection accuracy, and reduce the cost of the inspection apparatus.

  Further, in the automobile surface defect inspection system according to the present embodiment, two-dimensional image analysis and three-dimensional image analysis are performed at the same time. Any defect such as a dent without a change in color intensity can be detected relatively easily and with probability.

  INDUSTRIAL APPLICABILITY The present invention is useful as a surface inspection device, a surface inspection method, and a surface inspection program for measuring surface defects such as wrinkles, fading, paint peeling and dents on a measurement object, and inspecting dimensions, area, and volume in a non-contact manner . In particular, the present invention is suitable for surface inspection of a measurement object having a strong surface reflection, such as automobiles and parts thereof, metal products such as home appliances and porcelain products, and glossy products.

A Measurement object 1 Measurement space 1a Tent 1b Curved wall 2-1, 2-2, 2-3 Illumination device 3 Pattern light projection device 4 Camera device 5 Data processing device 6-1, 6-2, 6-3 6-4, 6-5 Transmission cable 10 Storage means 11 Illumination means 12 All illumination image imaging means 13 Projection pattern generation means 14 Pattern light projection means 15 Reflection pattern image imaging means 16 Three-dimensional image composition means 17 Two-dimensional image analysis means 18 Three-dimensional image analysis means 19 Defect detection classification size estimation means 20 Measurement result expression means 21 Storage output means

Claims (6)

A measurement environment configuration device that configures a measurement space having a curved wall formed of a translucent material;
The light source light provided outside the curved wall with respect to the measurement object in the measurement space, and the light source light that is transmitted through the curved wall and is totally irradiated to the measurement object is uniformly distributed. An illumination device for dimming and irradiating all ;
The pattern light is projected and reflected on the curved wall so as not to directly hit the measurement object, and the reflected pattern light (hereinafter referred to as “reflection pattern light”) is projected onto the measurement object. A pattern light projection device for
An imaging device for imaging the measurement object in a state of being stationary relative to the measurement object, the imaging and the reflected light pattern of the total illumination image of the measurement object, which is the total irradiation by the illumination device An imaging device for imaging the projected reflection pattern image of the measurement object;
A surface inspection apparatus including a data processing apparatus that detects a surface defect of the measurement object from the total illumination image and the reflection pattern image. The imaging device is plural,
The data processing device includes:
Three-dimensional image synthesis means for three-dimensionally synthesizing the entire illumination image and the reflection pattern image captured from each viewing angle by the plurality of imaging devices, and creating a wide range of inspection images;
Two-dimensional image analysis means for analyzing the color intensity of the entire illumination image and the reflection pattern image and detecting a region suspected of being a defect as a first defect candidate region;
Analyzing the synthesized image of the total illumination image and the reflection pattern image synthesized by the three-dimensional image synthesis means, calculating the depth coordinate value of the measurement point on the surface of the measurement object, and based on the depth coordinate value A three-dimensional image analysis means for detecting a region suspected of being a defect as a second defect candidate region;
Defect detection classification size estimation means for identifying and classifying defects in the first defect candidate area and the second defect candidate area;
The surface inspection apparatus according to claim 1 , comprising: The data processing device includes:
3. The measurement result expression means for displaying the defect detected by the defect detection classification size estimation means, marking a different color depending on the type of the defect, and displaying the estimated size of the defect. Surface inspection equipment.   The surface inspection apparatus according to claim 1, wherein the imaging apparatus captures the reflection pattern image from a viewing angle different from a projection angle of the projection pattern light. Configuring a measurement space having a curved wall formed of a translucent material;
The light source light provided outside the curved wall with respect to the measurement object in the measurement space, and the light source light that is transmitted through the curved wall and is totally irradiated to the measurement object is uniformly distributed. Dimming and full illumination ,
Taking a full illumination image of the measurement object that has been fully irradiated by an imaging device that images the measurement object in a state of being relatively stationary with respect to the measurement object;
The pattern light is projected and reflected on the curved wall so as not to directly hit the measurement object, and the reflected pattern light (hereinafter referred to as “reflection pattern light”) is projected onto the measurement object. about,
Imaging a reflection pattern image of the measurement object on which the reflection pattern light is projected, by an imaging device that images the measurement object in a stationary state relative to the measurement object;
A surface inspection method including detecting a surface defect of the measurement object from the total illumination image and the reflection pattern image. For a measurement object in a measurement space having a curved wall formed of a semi-transparent material, light source light that is transmitted through the curved wall and is totally irradiated to the measurement object is uniformly reduced. Full irradiation ,
Taking a full illumination image of the measurement object that has been fully irradiated by an imaging device that images the measurement object in a state of being relatively stationary with respect to the measurement object;
The pattern light is projected and reflected on the curved wall so as not to directly hit the measurement object, and the reflected pattern light (hereinafter referred to as “reflection pattern light”) is projected onto the measurement object. about,
Imaging a reflection pattern image of the measurement object on which the reflection pattern light is projected, by an imaging device that images the measurement object in a stationary state relative to the measurement object;
A surface inspection program for causing a computer to detect a surface defect of the measurement object from the entire illumination image and the reflection pattern image.

Images