JP5807772B2 - Defect detection apparatus and method - Google Patents

Description

  The present invention relates to a defect type determination apparatus and a determination method for optically detecting a defect generated on a surface of a workpiece after cutting such as a casting and determining the type of the defect. In particular, a defect detection device that captures surface images when light is irradiated on the inspection target surface of the workpiece from various directions, detects defects from feature amounts grasped from the pixel information, and determines the type thereof, and It relates to the determination method.

  In the cutting surface of aluminum castings, etc., indentation caused by gripping the workpiece with the cutting piece attached in the cutting process, subtle steps caused by the cutting sequence in the cutting process, variation in workpiece rigidity, etc. May cause chattering or the like due to blade vibration. These are classified into a non-defective part within the allowable tolerance and a defective part outside the allowable tolerance according to the depth, and what should be judged as a defective part must be able to be judged as a defect without omission.

  In order to automatically detect and determine such a defect, for example, a defect detection device that inspects a defect by irradiating light on a surface to be inspected of a workpiece, receiving the reflected light with a CCD camera, etc. It has been known. However, such a defect detection apparatus generally uses a smooth surface as an inspection target surface, and is not suitable for a defect on a rough surface.

  On the other hand, for example, in the defect detection apparatus described in Patent Document 1, there is a defect detection apparatus that can detect defects stably even on a surface where irregular cutting traces exist separately from defects. It is disclosed. The defect detection device disclosed in Patent Document 1 sets the intensity of the irradiation light so that the sensitivity of the light receiving element of the imaging means is saturated when the reflected light of the light irradiated to the part other than the defect is received, Defects are detected by changing the intensity of irradiation light under the set conditions.

  However, in the defect detection apparatus described in Patent Document 1, a defect having a certain depth can be detected. However, in the case of a defect such as chatter having a small surface roughness, the defect is detected in the same manner as a normal part. The sensitivity of the light receiving element to the reflected light is saturated. For this reason, there has been a problem that defects cannot be accurately detected.

  As a method for solving such a problem, a method of photographing a surface image from multiple directions by changing the light irradiation direction and moving the CCD camera in accordance with each direction can be considered. In other words, the defect portion has a feature that the overall reflection intensity is low with respect to a normal surface and the reflection intensity can change depending on the direction of irradiation light. By adopting a method of determining the type of defect by classifying the image information obtained from a plurality of photographed surface images according to the imaging direction by illuminating the inspection site from multiple directions by imaging The accuracy of defect detection can be increased.

Japanese Patent Application Laid-Open No. 2006-208259

  As a method of photographing the inspection surface by irradiating light from multiple directions, for example, a method of photographing the reflected light after fixing the direction of the irradiation light and changing the posture of the workpiece and the direction of the light receiving unit of the photographing means Can be considered. However, in the method of changing the posture of the work, the mechanism for adjusting the optical axis of illumination and the optical axis of reflected light tends to be complicated in an apparatus that realizes the method. In such an apparatus, it is conceivable to change the posture of the table to which the workpiece is attached to an actuator or the like. However, it is generally difficult to change the posture at high speed, and it takes time for defect inspection. Problems also arise.

  Further, when adjusting the optical axis of illumination and the optical axis of reflected light, it is difficult to adjust the distance of the optical path from the light emission position to the light receiving unit of the imaging means via the inspection surface (reflection surface). It is. Therefore, in order to compare a plurality of surface images photographed from multiple directions, each of the photographed surface images is a reference image photographed at a reference position (generally, the image is photographed using a horizontal state as a reference position). It is necessary to obtain a light distribution by converting to an image as a relative relationship to the (surface image). For example, in order to obtain the light distribution from a surface image taken at a certain position, the declination formed by the surface normal at that position and the surface normal at the reference position is obtained, and the normal part and defective part are determined. It is necessary to define a difference in the intensity of the reflected light as data with respect to the deviation angle, and to perform processing for conversion to a surface image photographed at the reference position by performing image conversion such that the deviation angle becomes zero. This requires a large calculation cost.

  In addition, when diffuse illumination by a wide-angle light-emitting element and diffuse projection / reception that receives light as it is are selected, in the obtained light distribution, the S / N ratio used for determination (for example, the reflection intensity and the normal part) The ratio with respect to the reflection intensity at the defect portion is low, which is disadvantageous in determining the type of defect. On the other hand, keeping the distance relationship constant according to the angle is troublesome in work and causes a problem that inspection takes time.

  For this problem, the direction of the inspection target surface is fixed with respect to the inspection area of the inspection target surface, the reflected light of the light irradiated from different directions is received, a surface image is taken, and the pixel according to the illumination direction This is solved by a defect detection apparatus and method for detecting defects from information.

  Further, without changing the direction of the surface of the inspection object, a light emitting means capable of irradiating the inspection area from different directions to the inspection area, and the light emitted by the light emitting means is focused and irradiated to the inspection area. Lens means and a light receiving portion fixed in a vertical direction with respect to the surface of the inspection area so as to face the inspection area, and receive the reflected light of the light emitted by the light emitting means in the inspection area Possible imaging means, and the imaging means receives the reflected light by the light receiving unit corresponding to the lighting direction of the light emitting means, and images an image of the surface of the inspection area corresponding to illumination from different directions As described above, the defect detection apparatus includes a control unit that controls the imaging unit.

  In addition, the step of emitting light from the light emitting means toward the inspection area from a different direction to the inspection area without changing the direction of the surface of the inspection object, and narrowing the light emitted by the light emitting means to the inspection area A step of irradiating light, and a step of photographing a surface image by receiving a reflected light of the light irradiated to the inspection area corresponding to lighting of the light emitting means with the light receiving unit facing the inspection area And a light distribution distribution as a reflected light intensity with respect to the lighting direction of the light emitting means based on the pixel value of the surface image photographed by the photographing means, and the light emitting means on the normal surface and the light emitting means The defect detection method includes a step of discriminating defects by comparing with a distribution of reflected light as reflected light intensity with respect to the lighting direction.

  According to the present invention, it becomes easy to select an optical system member for photographing a surface image to be inspected by irradiating light from multiple directions. Further, it is possible to simplify the mechanism for the light receiving unit of the photographing unit to receive the reflected light from multiple directions. Furthermore, it is possible to reduce the time required for defect detection.

  Further, according to the present invention, when obtaining a light distribution from a surface image to be inspected taken from multiple directions, a direct light distribution can be obtained without using a reference image such as a surface image in a horizontal state, Calculation cost can be reduced.

It is the figure explaining the principle of the defect discrimination | determination in the defect detection apparatus and its method of this invention. It is the figure which showed the defect detection apparatus which is Example 1 of this invention. It is the figure which showed the inspection head of the defect detection apparatus which is Example 1 of this invention. It is the figure which showed the outline of the whole structure including the control system of the defect detection apparatus which is Example 1 of this invention. It is the flowchart which showed the procedure of the defect detection apparatus of this invention. It is the surface image of the level | step difference of height 3 micrometers as an example of a non-defect part. It is the surface image of the level | step difference of height 8 micrometers as an example of a defect part. It is a surface image of the crimping | compression-bonding trace as an example of a defect part. It is a pseudo light distribution in the case of a step of 3 micrometers in height in parallel projection / reception. This is a pseudo light distribution in the case of a height difference of 3 micrometers in coaxial wide-angle light projecting and receiving. This is a pseudo light distribution in the case of a step of 8 micrometers in height in parallel projection / reception. This is a pseudo light distribution in the case of a height difference of 8 micrometers in coaxial wide-angle light projecting and receiving. It is a pseudo light distribution in the case of a crimp mark in parallel light projection and reception. It is a pseudo light distribution in the case of a crimp mark in coaxial wide angle light projecting and receiving. It is the figure which showed the defect detection apparatus which is Example 2 of this invention. It is the figure which showed the inspection head part of the defect detection apparatus seen from the arrow Y of FIG. It is the figure which showed the inspection head part of the defect detection apparatus seen from the arrow Z of FIG.

  With reference to FIG. 1, the principle of defect determination in the defect detection apparatus and method of the present invention will be described. FIG. 1 is a diagram showing the concept of an apparatus and method for optically detecting defects generated on the surface of a workpiece (inspection object), for example. The surface to be inspected of the inspection object is defined as the inspection target surface, and the region to be inspected of the inspection target surface is defined as the inspection region W. In addition, in the inspection target area W, in particular, a range in which defects are to be detected in one sequence is defined as an inspection area T. The defect detection apparatus and method of the present invention include a light emitting means, a lens means, a photographing means, and a control unit.

The light emitting means has a curved orbit O centered on the inspection area T of the surface to be inspected in the inspection area of the surface of the inspection object without changing the direction of the posture of the inspection object (surface of the inspection object). Irradiation is possible toward the examination area T from different directions on _I (for example, positions O ₁ , O ₂ , O ₃ , O ₄ , O ₅ , O ₆ , O ₇ or O ₈ ). As the light emitting means, for example, a plurality of light emitting elements fixed at arbitrary positions on the track O _I can be used. In addition, one light-emitting element or a small number of light-emitting elements are moved on the orbit O _I , and each position (for example, positions O ₁ , O ₂ , O ₃ , O ₄ , O ₅ , O ₆ , O _7, or O _8). ) Can also be turned on.

The lens means squeezes the light emitted by the light emitting means and irradiates the inspection area T. The light irradiated to the examination area T does not need to be a parallel light beam. However, if a collimator lens is used as the lens means and the light emitted from the light emitting means can be collimated to irradiate the inspection area T, scattering in the inspection area T can be prevented and more accurate defect determination can be performed. . The lens means is arranged corresponding to the position of the light emitting element. For example, in the case of a plurality of fixed light emitting elements, a plurality of lenses L ₁ , L ₂ , L ₃ , L ₄ , L ₅ , L ₆ , L ₇ or L ₈ may be arranged. Or, in the case of moving the light-emitting element in orbit O _I, go with it, always may be located at each optical path between the examination zone T from the respective light emitting elements. Furthermore, even when the light emitting element is moved on the track O _I , a plurality of lenses L ₁ , L ₂ , L ₃ , L ₄ , L ₅ , L ₆ , L ₇ or L ₈ are arranged. Thus, the light emitting element moved at each position where the lens is disposed may be turned on.

  The imaging means C is arranged at a position where the light receiving portion receives the reflected light of the inspection area T and the surface of the inspection area T can be imaged. The photographing means C is preferably arranged so as to be fixed in the vertical direction with respect to the surface of the inspection area T so as to face the inspection area T. Thereby, in measuring the reflection intensity, it is possible to perform comparison processing as reflected light that is symmetric in all directions from the vertical direction with respect to the surface of the inspection area T.

The control unit (not shown in FIG. 1) controls the photographing unit C in directions (O ₁ , O ₂ , O ₃ , O ₄ , O ₅ , O ₆ ) in which the photographing unit C is different according to the lighting direction of the light emitting unit. , O ₇ or O ₈ ) is controlled so as to take an image of the surface of the examination area T corresponding to the illumination from the light source. For example, when the light emitting means is a plurality of light emitting elements fixed at arbitrary positions on the track O _I , the photographing means C is controlled to perform photographing at the lighting timing of each light emitting element. Further, in the case of lighting one or a small number of light emitting elements are moved in orbit O _I, the photographing means C at the timing of the lighting at each position is controlled to perform photographing. In detecting and discriminating defects, it is necessary to determine in what direction the captured image is the image captured (pixel information obtained) when illuminated. Therefore, when the surface image is captured, information that associates the captured image with the illumination direction is generated. The illumination direction can be expressed, for example, as an angle from the vertical direction with respect to the surface of the inspection area T, or the lighting position of the light emitting means as an angle from the horizontal direction.

When the light emitting means is turned on from different directions (O ₁ , O ₂ , O ₃ , O ₄ , O ₅ , O ₆ , O _7, or O ₈ ) and irradiated to the inspection area T, irradiation is performed on a normal surface without defects. The light generates uniform reflected light on the surface of the inspection area T by regular reflection. That is, it reflects so that the incident angle and the reflection angle are the same with respect to the surface of the inspection area T.

On the other hand, when a defect exists, the surface has a micro irregular surface. Therefore, when the light emitting means is turned on from different directions (O ₁ , O ₂ , O ₃ , O ₄ , O ₅ , O ₆ , O _7, or O ₈ ) and irradiates the inspection area T, it is irregular in either direction. Produces reflected light. That is, the incident angle and the reflection angle are not the same with respect to the surface of the inspection area T. Therefore, when the intensity of the reflected light on the normal surface and the intensity of the reflected light on the surface having a defect are compared as a light distribution, a characteristic difference occurs when the defect is present. That is, for example, when the light emitting means is irradiated onto the inspection area T from the direction of O ₁ , if there is no defect on the surface of the inspection area T, the light is reflected from the opposite light emitting means toward the direction of O _8. Therefore, the photographing means C does not receive the reflected light R. However, at this time, if the photographing means C receives the reflected light R, the direction of the reflected light with respect to the irradiation direction is not the same as the incident angle and the reflected angle with respect to the surface of the inspection area T. Since it can be determined, it can be understood that some defect exists in the inspection area T. Also, different characteristics are seen in the reflected light intensity depending on the type of defect. When parallel light projection or parallel light reception is performed, the difference in reflected light intensity appears more remarkably.

Assuming a hemisphere centered on the surface of the inspection area T, the light emitting means is perpendicular to the surface of the inspection area T when viewed in a plane perpendicular to the surface of the inspection area T through the zenith of the hemisphere. To both directions so that it can be photographed at an arbitrary angle up to 90 degrees (horizontal). It should be noted that at least defects can be detected as long as the image can be taken at an angle from the vertical direction to 60 degrees in both directions with respect to the surface of the inspection area T. Furthermore, as the O _II of Figure 1, any one surface perpendicular to the surface of the examination zone T is in about a vertical direction with respect to the surface of the examination zone T, set in all directions (360 degrees) The light emitting means are attached on these multiple surfaces. As a result, it becomes possible to detect defects grown in a biased direction.

(Example 1)
The defect detection apparatus 1 according to the first embodiment of the present invention will be described with reference to FIGS. Example 1 corresponds to a mode in which a plurality of fixed light emitting elements are applied to arbitrary positions on the omnidirectional trajectory O _{I in the} above-described embodiment. FIG. 2 is a diagram illustrating an appearance of the defect detection apparatus 1. FIG. 3 is a diagram showing an outline of a cross section of the inspection head 2 shown by a cross section XX in FIG. FIG. 4 is a diagram showing an outline of a configuration of a control system for operating the defect detection apparatus 1. The defect detection apparatus 1 includes an inspection head 2 and an imaging control unit 3.

  In FIG. 2, the workpiece 5 to be inspected is attached to the workpiece support 6, and the workpiece 5 is arranged with the inspection target surface 5 a to be checked for defects on the surface of the workpiece 5 facing the inspection head 2. To do. Here, it defines as inspection area 5a among inspection object surfaces. Therefore, the defect inspection means that the entire inspection area in the inspection target surface is inspected while shifting the position of the inspection area 5a. The inspection area 5 a is set so as to be arranged at a predetermined position corresponding to the lower center of the inspection head 2.

  The inspection head 2 includes a plurality of light emitting elements 21, a lens 22, and an imaging unit 23. The inspection head 2 is attached to the attitude control device. Each of the plurality of light emitting elements 21 has a different direction with respect to the inspection area 5a when the inspection head 2 is arranged such that the inspection area 5a is positioned at a predetermined position. Placed in. In addition, the plurality of light emitting elements 21 are also arranged so as to be located in different directions with respect to the direction from the inspection area 5 a to the imaging unit 23. Specifically, when one point of the inspection area 5a is set as an origin (generally, approximately the center of the inspection area 5a), the optical axes of the plurality of light emitting elements 21 toward the origin are set to face each other. Means that.

  FIG. 3 shows a cross section of the inspection head 2 taken along a section XX in FIG. 2 as a specific example of the inspection head 2. As shown here, for example, a plurality of light emitting elements 21 are arranged on a spherical cover member 24. In particular, the cover member 24 is preferably a hemisphere. The inspection area 5a is arranged at the center of the opening bottom of the spherical cover member 24 (in the case of a hemisphere, the center of the hemisphere). A plurality of light emitting elements 21 can also be arranged on a polyhedral cover member or a truss member in which each side of the polyhedron is configured by a frame. Even if a hemisphere is used, the predetermined position corresponding to the center of the inspection head 2 where the inspection area 5 a is to be arranged does not have to be constant from each of the light emitting elements 21. However, in the case of diffused light, each of the plurality of light emitting elements 21 is disposed at a substantially constant distance from a predetermined position corresponding to the center of the inspection head 2 where the inspection area 5a is scheduled to be disposed. The accuracy may increase.

  Assuming a hemisphere centered on the surface of the inspection area 5a, the plurality of light emitting elements 21 are arranged in a line in any one plane passing through the zenith of the hemisphere and perpendicular to the surface of the inspection area T. It is preferable that they are arranged. Furthermore, it is preferable that the plurality of light emitting elements 21 are arranged so that a plurality of light emitting elements 21 form a plurality of rows at arbitrary angles in all directions (360 degrees) with the photographing unit 23 as the center.

  A lens 22 is disposed between the light emitting element 21 and the inspection area 5a. As long as the plurality of light emitting elements 21 can irradiate the inspection area 5a, any light source can be used. However, if a parallel light beam can be irradiated, a particularly remarkable advantageous effect is obtained. In the present embodiment, the lens 22 is disposed between the light emitting element 21 and the inspection area 5a, and can irradiate the inspection area 5a with parallel rays. As the posture control device, typically, an articulated robot arm 25 can be used. The light-emitting element 21 is typically a light-emitting diode (hereinafter referred to as an LED), but other illumination members can be used as long as they can be converted into parallel rays by combination with the lens 22. The imaging unit 23 is arranged so that the light receiving unit 23a of the imaging unit 23 faces the position set so that the inspection area 5a is arranged, and the light receiving unit 23a is positioned in the vertical direction with respect to the surface of the inspection area 5a. . The light receiving element portion 23b of the photographing means 23 can be, for example, a CCD or CMOS high-speed camera. Each of the plurality of light emitting elements 21 reflects the light E emitted from multiple directions toward the inspection area 5a at the inspection area 5a, and the imaging unit 23 receives the parallel reflected light R from the inspection area 5a.

  As shown in FIG. 4, the articulated robot arm 25 is connected to the robot control device 34. The robot controller 34 is configured such that the light receiving portion 23a of the imaging means 23 of the inspection head 2 attached to the tip faces the inspection area 5a, and the light receiving portion 23a is positioned in the vertical direction with respect to the surface of the inspection area 5a. The joint of the multi-joint robot arm 25 is moved.

  The workpiece support 6 includes, for example, an X table 26 movable in one horizontal direction and a Y table 27 movable in a direction orthogonal to the movable direction of the X table 26. Furthermore, a rotary table 28 can be added. The X table 26, the Y table 27, and the rotary table 28 of the work support 6 are connected to the stage control device 35, and move the work 5 in two directions orthogonal to each other in the horizontal direction or rotate it in the horizontal plane. It is possible.

The imaging control unit 3 includes a control unit 32 and a lighting device 33. The plurality of light emitting elements 21 are connected to the lighting device 33, and are turned on and off by receiving voltage supply from the lighting device 33. The control unit 32 is connected to the robot control device 34. The control unit 32 can transmit a joint operation command of the articulated robot arm 25 to the articulated robot arm 25 and can receive information on the state of the articulated robot arm 25.
The illumination device 33 is connected to the control unit 32. The control unit 32 can control the lighting and extinguishing timing of each of the plurality of light emitting elements 21 and the luminance of the light emitting element 21 by adjusting the voltage to the plurality of light emitting elements 21 to be supplied to the lighting device 33. For example, all of the plurality of light emitting elements 21 are turned off, and first, the one light emitting element 21 located at the end is turned on. The light emitting element 21 adjacent to the light emitting element is turned on at the timing when the light emitting element 21 is turned off. This is repeated sequentially. The lighting time can be, for example, 1 millisecond. Accordingly, the plurality of light emitting elements 21 arranged in a row are repeatedly turned on and off in sequence. The lighting / extinguishing of the plurality of light emitting elements 21 is not necessarily performed in the order of the adjacent light emitting elements 21, and as long as the photographing unit 23 performs imaging according to the position of the light emitting elements 21 that are lit, the inspection area What is necessary is just to be linked | related as a surface image of the test | inspection area 5a with respect to the angle which the imaging | photography means 22 and the light emitting element 21 make with 5a. That is, the control unit 32 controls the photographing unit 23 to photograph an image of the surface of the examination area 5a according to the position of the light emitting unit 21 that is lit among the plurality of light emitting elements 21, and the photographed surface image and Information that associates the position of the light-emitting element that is lit is generated. The generated information can be stored in a storage device (not shown) or the like. As for the position of the light-emitting element that has been turned on, for example, an arbitrary one direction in the inspection target surface (horizontal plane) is defined as the X axis (for example, the movable direction of the X table 26 of the work support 6), and the direction orthogonal thereto is defined as Y. An axis (for example, the moving direction of the Y table 27 of the work support 6), the vertical direction with respect to the surface of the horizontal inspection area 5a is assumed to be the Z axis, and the counterclockwise direction from the X axis to the Z axis is positive. The polar angle (θ, φ) can be expressed by the declination angle θ in this direction and the declination angle φ from the Z axis. In the case of Example 1, since all of the plurality of light emitting elements 21 are fixed, the deflection angle θ and the deflection angle φ are uniquely assigned to each light emitting element. Since it is possible to irradiate light from different directions without changing the direction of the surface of the inspection object, common polar coordinates with respect to the surface of the inspection area 5a without conversion regardless of the direction of illumination. It can be expressed as (θ, φ).

  In the case where the light emitting element 21 is an LED, it is possible to reduce the time from the turn-off state to the turn-on state until the light is turned off again in a short time of several hundred microseconds to several milliseconds. Thereby, the time from the lighting of the first element of the plurality of light emitting elements 21 in a row to the extinguishing of the last element can be extremely shortened.

  The imaging unit 23 is also connected to the control unit 32 and adjusts the timing at which the imaging unit 23 captures the reflected light. The photographing means 23 performs photographing in accordance with the lighting timing of the plurality of light emitting elements 21. The captured video is also stored in the control unit 32. If the light emitting element 21 is an LED and the photographing means 23 is a high-speed CMOS camera, as a result of imaging in accordance with the timing of light emission of the LED, the light irradiated from multiple directions in about 1 second with respect to a visual field of 50 square millimeters. Shooting of multiple corresponding surface images can be completed.

  The control unit 32 is connected to the image forming apparatus 31. The surface image captured by the imaging unit 23 and acquired by the control unit 32 is sent to the image forming apparatus 31. The surface image is temporarily stored in a memory (not shown). The image forming apparatus 31 combines all pixel data of the surface image corresponding to the stored light irradiated from multiple directions, and calculates the pseudo light distribution of the entire inspection region. The pseudo light distribution is a two-dimensional representation of luminance (reflected light intensity) as a vector quantity and a deviation angle φ in the direction of reflected light from the inspection area 5a as a vector element in a two-dimensional plane of an arbitrary deviation angle θ. It is intensity distribution of the reflected light on a dimension surface. From the calculated pseudo light distribution, feature amounts of all pixels in the inspection area are calculated. The feature amount is defined as F (I, φ) from the pixel value of the photographed surface image by the reflected light intensity (luminance) I and the deflection angle φ at that time. Since the plurality of light emitting elements 21 are arranged in all directions (360 degrees) with the photographing unit 23 as the center, when there is a defect, a difference in reflection characteristics appears in any direction. Further, the difference in the reflection characteristics varies depending on the type of defect. Therefore, if this is grasped in advance as the feature amount F (I, φ) of each surface at the deflection angle θ (θ = 0 to 360 °), the type of defect can be grasped by comparing the feature amounts. it can.

Here, in the present embodiment, in the defect detection and the type determination, between the reflected light when the light is irradiated for photographing the surface image and the reflected light when the normal portion is irradiated with the light in advance. Irradiation of “measured reflected light intensity” with respect to “reflected light intensity acquired in the normal part in advance” as an index of reflection characteristic difference (exceeding measurement tolerance) in the pseudo light distribution It discriminate | determines by the value corresponding to the direction of the light (angle with respect to the normal line of a test object surface). Specifically, the feature quantity F _N (In, φn) (−90 ° <n <90 °) is obtained in the pseudo light distribution on the normal surface at an arbitrary deflection angle θ. On the other hand, from the pseudo light distribution obtained by the captured surface image for defect inspection at the same deflection angle theta, the feature amount _{F M (Im, φm) (} - 90 ° <n <90 °) obtained. Request deflection angle characteristic of the ratio when the feature amount F _M feature amount F _N and the same deflection angle φm (= φn) (Im / In) is maximized. The feature value ratio for each defect type was previously acquired as a threshold value TH, and the feature value ratio obtained from the measured image was compared with the threshold value, and the measured feature value ratio exceeded TH. In some cases, it can be recognized as a defect.

  Note that the presence / absence of a defect can be determined by another method instead of the above method. The height of the defect can be recognized by the brightness of the reflected light represented by the pixel value of the photographed image. Therefore, whether or not the binarization data of the brightness corresponding to the threshold value TH is determined in advance, and the brightness for all of the pixel values of the photographed image is greater than the binarization data of the threshold brightness Determine if it is small. Then, in each pixel, if it is larger than the binarized data of the threshold brightness, a binary value of the lowest brightness is given, and if it is smaller, a binary value of the highest brightness is given. As a result, an image in which a pixel corresponding to black is given to a pixel where a specific difference is seen in the reflected light, for example, and a pixel corresponding to white is given to a location where no specific difference is seen, for example. The size of the defect can be determined by the spread of a specific difference. Therefore, if this is performed on the pixel values of all the images photographed by applying light rays from different directions and synthesized, for example, it is possible to determine that a portion having a large black pixel has a defect.

  Next, the defect detection procedure will be described with reference to FIG. First, a work is placed on the work support 6. Then, the X table 26, the Y table 27, or the rotary table 28 of the work support 6 is rotated so that the inspection area 5a is positioned at a predetermined position of the inspection head 2, and the joint of the articulated robot arm 25 is operated ( S1). The controller 32 adjusts the voltage supplied from the lighting device 33 to turn on and off each of the plurality of light emitting elements 21 (S2), and take a surface image when each of the light emitting elements 21 is turned on ( S3). The luminance value of all the pixel values as the surface image of the inspection area 5a is stored in the memory in association with the information on the angle formed by the imaging means 22 and the light emitting element 21 with respect to the inspection area 5a ( S4). Thereafter, it is determined whether or not imaging at all designated angles corresponding to the inspection area has been completed. If the imaging has not been completed, the imaging process (S2, S3) is performed again on all of the plurality of light emitting elements 21. , S4) is repeated (S5). If completed, the pseudo light distribution of all pixels in the inspection area is calculated (S6). Subsequently, after the pseudo light distribution is obtained for all the pixels of all photographed surface images, the feature amount is calculated (S7). Thereafter, the normal surface and the defect type are determined from the obtained feature amount (S8). Pixels are classified for each type of defect (S9), a feature amount is calculated for each type, and compared with a threshold value. When the determination of all inspection areas is completed (S11), the presence or absence of defects is determined for those exceeding the threshold (S12).

  Here, in the optical system of the present embodiment, referring to FIGS. 6 to 8, FIGS. 9 and 10, FIGS. 11 and 12, and FIGS. 13 and 14, the light emitting elements 21 and the lenses 22 are parallel. How the light projection and the parallel light reception by the photographing means 23 have an advantageous effect will be described. As an example, a non-defective part has a step (height 3 micrometers) as in the surface image shown in FIG. 6, and a defective part as in the surface image shown in FIG. 7 as an example of the defective part. Consider the case of (having a height of 8 micrometers) and the case of having a crimp mark like the surface image shown in FIG.

  In the case of defect detection on the surface having the step (height 3 micrometers) shown in FIG. 6, the pseudo arrangement measured and calculated by the defect detection apparatus of this embodiment as the inspection area shown by a circle in FIG. The light distribution is shown in FIG. In this pseudo light distribution, the reflected light intensity 51 acquired in the normal part in advance and the reflected light intensity 52 of the stepped part measured for defect detection are overlapped. In this example, the distribution of the reflected light intensity 51 (vector amount) acquired in the normal portion in advance shows a substantially uniform fan shape in all directions, and the reflected light intensity 52 (vector amount) of the step portion is Although the size is slightly lower, it can be seen that the distribution is uniform. Further, the ratio of the maximum value a of the reflected light intensity 52 (vector amount) of the measured stepped portion to the maximum value b of the reflected light intensity 51 (vector amount) acquired in advance in the normal part (hereinafter referred to as “light distribution distribution ratio ( a / b) ”) is 0.71.

  In the case of detecting a defect on the surface having the step (height 8 micrometers) shown in FIG. 7, the pseudo arrangement measured and calculated by the defect detection apparatus of this embodiment as the inspection area shown by a circle in FIG. The light distribution is shown in FIG. As in this example that is determined to be a defect, in the simulated light distribution of the step having a large height, the reflected light intensity 51 (vector amount) acquired in the normal portion in advance showing an even distribution is higher than the measured step portion. It can be clearly seen that the reflected light intensity 52 (vector quantity) has a strong bias especially in one direction (left side in FIG. 11). In addition, the light distribution ratio (a / b) is a high value of 6.6.

  In the case of defect detection on the surface having the crimp mark shown in FIG. 8, the pseudo light distribution measured and calculated by the defect detection apparatus of the present embodiment as the inspection area indicated by a circle in FIG. 8 is shown in FIG. Show. In this example, the reflected light intensity 52 (vector amount) of the measured stepped portion is particularly in one direction (the right side in FIG. 13) rather than the reflected light intensity 51 (vector amount) acquired in the normal part in advance showing an even distribution. It can be seen that there is a very strong bias. Further, it is clear that this bias tends to be different from the pseudo light distribution in the case of a step (height of 8 micrometers). The light distribution ratio (a / b) in this case is 1.48, which is lower than that in the case of a step (height 8 micrometers), but is higher than that in a non-defect portion.

  As described above, in the optical system according to the present embodiment, parallel light projection by the plurality of light emitting elements 21 and the lenses 22 and parallel light reception by the photographing unit 23 can be realized. In the present embodiment, by realizing parallel light projection / reception, the sensitivity of the light distribution ratio (a / b) can be increased to 6.6 in the case of a crimp mark. This is an extremely high value as compared with the case where a coaxial wide-angle light projecting optical system combining a coaxial surface illumination and a normal lens and a coaxial wide-angle light projecting / receiving optical system that receives the reflected light are used. That is, in the determination of the presence / absence of a defect portion and the classification of the type of defect portion, it means that the sensitivity to defects is high and the determination is easy.

  10, 12 and 14 are pseudo light distributions in the case of coaxial wide-angle light projection and reception, respectively. As shown in FIG. 10, FIG. 12, and FIG. 14, in the case of the coaxial wide-angle projecting / receiving optical system, the height of 3 μm, the step of 8 μm, and the crimp mark are pre- Based on the obtained pseudo light distribution of the reflected light intensity 51, the shape of the measured pseudo light distribution of the reflected light intensity 52, and the vector quantity of the reflected light intensity in each direction, the pseudo light distribution in the case of the parallel light receiving and receiving optical system is used. There is no significant difference as much as the light distribution. In particular, in the case of a coaxial wide angle light projecting / receiving optical system, the shape of the pseudo light distribution (step 12) having a height of 8 micrometers to be determined as a defect and the crimp mark (FIG. 14) is large. It can be seen that there is no difference and it is difficult to determine the type of defect.

  Further, in the pseudo light distribution (FIG. 6) in the case of a step having a height of 3 micrometers that does not become a defect, the light distribution ratio (a / b) in the parallel light projecting / receiving optical system is 0.71 (FIG. 9). On the other hand, the light distribution ratio (a / b) in the coaxial wide-angle projecting / receiving optical system is substantially close to 1.1 (FIG. 10). In the case of the pseudo light distribution (FIG. 7) in the case of a height difference of 8 micrometers recognized as a defect, the light distribution ratio (a / b) in the parallel light projecting / receiving optical system is 6.6 (FIG. 11) and a large change with a higher sensitivity than in the case of a height difference of 3 micrometers. However, the light distribution ratio (a / b) in the coaxial wide-angle projecting / receiving optical system is 1.6 (FIG. 12), which is 1.1 (which is a light distribution ratio (a / b) of a step of 3 micrometers in height. The change is small with respect to FIG. That is, the light distribution distribution ratio (a / b) obtained by the present invention using the parallel light projecting / receiving optical system has a very high sensitivity to the defect, so that the feature comparison by the defect can be performed with a high sensitivity. It is possible to detect and judge defects with high accuracy.

(Example 2)
Next, the defect detection apparatus 1 according to the second embodiment of the present invention will be described with reference to FIGS. 15, 16, and 17. Example 2 corresponds to a mode in which the light emitting element is moved along the omnidirectional trajectory O _I in the embodiment. Example 2 is the same as Example 1 except that the light-emitting element is moved to irradiate the inspection area 5a with light from different directions. Here, a different part from Example 1 is demonstrated. FIG. 15 is a diagram illustrating an appearance of the defect detection apparatus 1 according to the second embodiment. 16 is a view as seen from an arrow Y in FIG. 15, and FIG. 17 is a view as seen from an arrow Z. The configuration of the control system for operating the defect detection apparatus 1 is the same as that shown in FIG. The defect detection apparatus 1 according to the second embodiment also includes an inspection head 2 and an imaging control unit 3. The work 5 is also attached to the work support 6 in the same manner as in the first embodiment, and the work support 6 also operates in the same manner as in the first embodiment.

  The light receiving unit 23a of the imaging unit 23 is disposed in the vertical direction from the inspection area 5a with the light receiving portion 23a facing the surface of the inspection area 5a of the workpiece 5. The inspection head 8 according to the second embodiment includes a light emitting element 41, a lens 42, and a half mirror 43. The half mirror 43 is disposed so as to be inclined between the optical path between the inspection area 5 a and the light receiving unit 23 a of the imaging unit 23. The inclination angle is arranged with an inclination of approximately 45 degrees with respect to the optical path between the inspection area 5a and the light receiving portion 23a of the imaging means 23. The light emitting element 41 is disposed so as to be orthogonal to the optical axis of the optical path between the inspection area 5 a and the light receiving unit 23 a of the imaging unit 23. The inspection head 8 is attached to a rotating means so that the half mirror 43 can rotate on the track around the inspection area 5a. More specifically, the half mirror 43 can rotate on an arcuate track that can move on the inspection area 5a perpendicular to the surface of the inspection area 5a. As the rotating means, for example, the joint robot arm 25 can be used, but various types of drive mechanisms are applicable as long as the half mirror 43 can rotate on a surface perpendicular to the surface of the inspection area 5a. it can. Even if the rotating means is rotated, the half mirror 43 is 45 degrees with respect to the optical path at each position in the optical path from the surface of the inspection area 5a to the light receiving unit 23a of the photographing means 23 at all positions. Located to keep the slope of. The rotating trajectory does not have to be a surface perpendicular to the surface of the inspection area 5a. However, if the surface is perpendicular to the surface of the inspection area 5a, it is possible to make a particularly accurate determination.

  The rotating means rotates the light emitting element 41 around the inspection area 5a and rotates the work 360 degrees by the rotary table 28, so that the inspection area 5a can be irradiated with light from all directions. The lens 42 is a collimator lens disposed between the light emitting element 41 and the half mirror 43, and irradiates parallel light to the half mirror 43. The light emitting element 41 is arranged so as to be able to irradiate light toward the half mirror 43 at all positions on the orbit where the light emitting element 41 rotates by the rotating means.

When light is emitted toward the half mirror 43 at an arbitrary position on the orbit where the light emitting element 41 rotates, the parallel light emitted from the light emitting element 41 is directed 90 degrees toward the inspection area 5a by the half mirror 43. The light is reflected and irradiated to the inspection area 5a. Of the reflected light reflected by the surface of the inspection area 5 a, a part of the reflected light travels straight in the vertical direction with respect to the surface of the inspection area 5 a, passes through the half mirror 43, and enters the light receiving unit 23 a of the imaging unit 23. As a result, the examination area 5a is illuminated from multiple directions, and the reflected light reflected there is received by the imaging means 32, and a surface image can be taken for illumination from different directions.

  Note that, in the same manner as in the first embodiment, the light distribution is obtained based on the pixels of the photographed surface image, and the defect is detected and discriminated from the feature amount ratio.

DESCRIPTION OF SYMBOLS 1 Defect detection apparatus 2, 8 Detection head 5 Work 21, 41 Light emitting element 23 Imaging means

Claims (11)

A defect detection device for detecting and judging defects generated on the surface of an inspection object,
A light emitting means capable of irradiating the inspection area from different directions toward the inspection area;
Lens means for squeezing the light emitted by the light emitting means and irradiating the inspection area;
An imaging means that is fixed in a vertical direction with respect to the surface of the inspection area so that a light receiving portion faces the inspection area, and that can receive the reflected light of the light emitted by the light emitting means in the inspection area. ,
The imaging means controls the imaging means so that the reflected light is received by the light receiving unit in response to lighting of the light emitting means, and the surface image of the examination area corresponding to illumination from different directions is taken. A control unit ;
An image processing unit connected to the photographing unit and the control unit,
The image processing unit generates a photographing light distribution for each pixel of the surface image as a reflected light intensity with respect to a lighting direction of the light emitting unit based on a value of a pixel of the surface image photographed by the photographing unit, A defect detection apparatus that determines a defect by comparing the photographing light distribution with a light distribution for each pixel of an image of reflected light as reflected light intensity with respect to a lighting direction of the light emitting unit on a normal surface . The defect detection apparatus according to claim 1,
The defect detection apparatus according to claim 1, wherein the lens unit is capable of irradiating the light emitted from the light emitting unit as a parallel light beam on the inspection area. The defect detection apparatus according to claim 1 or 2,
The defect detection apparatus, wherein the defect is determined by comparing a ratio between a reflected light intensity in the photographing light distribution and a reflected light intensity in the light distribution on a normal surface with a predetermined threshold . The defect detection apparatus according to any one of claims 1 to 3,
The light emitting means is a plurality of light emitting elements,
The lens means is a plurality of collimator lenses,
Each of the plurality of light emitting elements is disposed in a different direction from the inspection area of the surface,
Each of the plurality of collimator lenses can irradiate the inspection area with light emitted by each of the plurality of light emitting elements,
The control unit sequentially turns on and off each of the plurality of light emitting elements, and controls the photographing unit to photograph an image of the surface of the examination area according to the position of the light emitting unit that is lit. A defect detection apparatus that generates information associating a displayed image with a light-emitting element that is lit. The defect detection apparatus according to claim 4 ,
The defect detection apparatus , wherein the plurality of light emitting elements are mounted on a spherical surface . The defect detection apparatus according to claim 4 ,
The defect detection apparatus, wherein the plurality of light emitting elements are attached to members constituting a polyhedron . The defect detection apparatus according to any one of claims 1 to 3 ,
The light emitting means is attached to a rotating means that enables the light emitting means to rotate about the inspection area,
The defect detection apparatus , wherein the control unit captures an image of a surface of the inspection area according to a position of the light emitting unit, and generates information in which the captured image is associated with the position . The defect detection apparatus according to any one of claims 4 to 7 ,
Each of the light emitting elements is a light emitting diode element . A defect detection method for detecting and determining defects generated on the surface of an inspection object,
Emitting light from the light emitting means toward the inspection area from different directions to the inspection area on the surface of the inspection object;
Squeezing the light emitted by the light emitting means and irradiating the inspection area with light; and
Directly facing the inspection area to the inspection area, receiving reflected light of the light irradiated to the inspection area corresponding to the lighting of the light emitting means, and photographing a surface image;
Based on the pixel value of the surface image photographed by the photographing means, a photographing light distribution for each pixel of the surface image as a reflected light intensity with respect to the lighting direction of the light emitting means is generated, and the photographing light distribution distribution and And a step of discriminating a defect by comparing with a light distribution for each pixel of an image of reflected light as reflected light intensity with respect to a lighting direction of the light emitting means on a normal surface . The defect detection method according to claim 9 ,
The defect detection method according to claim 1, wherein the lens unit is capable of irradiating light emitted from the light-emitting unit as parallel light to the inspection area . The defect detection method according to any one of claims 9 to 10,
The step of determining the defect includes a step of determining a ratio between a reflected light intensity in the photographing light distribution and a reflected light intensity in the light distribution on a normal surface by comparing with a predetermined threshold value. Feature defect detection method.

Images