JP2010066126A - Image processing method and foreign matter inspecting device using the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing method for enhancing the recognition precision of a capture image in the inspection of the foreign matter in a groove. <P>SOLUTION: The image processing method includes measurement step (S201) for measuring the brightness of an image at every groove, gain calculation step (S202) for calculating a gain correcting the brightness of the image of a predetermined groove by using the average brightness of the brightnesses of the images of the grooves positioned on both sides of the predetermined groove, and gain correction step (S203) for correcting the brightness of the image of the predetermined groove at every groove by using the calculated gain. In rectilinear grooves arranged at regular intervals, the brightnesses of the images of the grooves positioned on both sides of the predetermined groove have linear relationship with respect to the brightness of the image of the predetermined groove. Accordingly, the brightness of the image of the predetermined groove not affected by an inspection environment can be estimated. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing method and a foreign matter inspection apparatus using the same, and more particularly to an image processing method for inspecting the presence or absence of foreign matter in a groove of an object to be inspected and a foreign matter inspection apparatus using the image processing method.

Conventionally, in foreign matter inspection that detects foreign matter in the deep groove that divides each segment of the commutator of the fuel pump, it is difficult to detect the presence or absence of foreign matter in the deep groove by capturing an image in the deep groove from the upper side in the axial direction of the commutator Met. On the other hand, a technique for inspecting the appearance of an object to be inspected from various angles using a prism is known. However, when this technology is applied to the inspection of foreign matter in the deep groove of a commutator and the inside of the deep groove of the commutator is inspected by a conical prism, the misalignment between the central axis of the conical prism and the central axis of the commutator, the inclination of the conical prism, and Due to the attenuation of light due to the composition of the conical prism, etc., there has been a problem that variation in density occurs in the captured image data, making it difficult to detect foreign matter. If the foreign matter in the deep groove is not detected, there is a possibility that the adjacent segments are conductive by the foreign matter.
By the way, in Patent Document 1, the luminance gradation of the video signal is corrected based on correction data set in advance for each luminance gradation distribution pattern.
However, in the inspection of foreign matter in the deep groove of the commutator, since there is no regularity in the luminance change of the measurement site in the captured image data, it is difficult to improve the recognition accuracy of the captured image by applying such a correction method. there were.

JP-A-6-313191

  An object of the present invention is to provide an image processing method for improving the recognition accuracy of a captured image in the inspection of foreign matter in a groove and a foreign matter inspection apparatus using the image processing method.

  According to the first aspect of the present invention, an image processing method for measuring the presence / absence of a foreign substance in a groove of an object to be inspected includes a measurement stage for measuring the luminance of an image for each groove, and a groove located on both sides of a predetermined groove A gain calculating step for calculating a gain for correcting the luminance of the image of the predetermined groove using the luminance obtained by averaging the luminance of the image, and a gain correction step for correcting the luminance for each image of the predetermined groove using the gain And including. In the linear grooves arranged at regular intervals, the luminance of the image of the groove located on both sides of the predetermined groove has a linear relationship with the luminance of the image of the predetermined groove. For this reason, the gain for correcting the luminance of the image of the predetermined groove is calculated by using the luminance obtained by averaging the luminance of the images of the grooves located on both sides of the predetermined groove, so that the inspection environment such as illumination variation can be calculated. The brightness of each groove that is not affected can be estimated. Thereby, the recognition accuracy of the captured image can be improved. As a result, a precise foreign object inspection can be performed. The inspected object of claim 1 may be inspected objects having various shapes such as a rectangular parallelepiped shape and a radial shape.

  According to a second aspect of the present invention, the image processing method further includes an average value adjusting step of correcting the luminance of the image for each groove and adjusting the average value of the luminance of the image for each groove to a predetermined value. For this reason, in the gain correction stage, the luminance of the image for each groove is relatively corrected based on a predetermined value. This increases the threshold margin in the binarization process. As a result, it is possible to improve the recognition accuracy of the captured image and perform a precise foreign object inspection.

  According to the invention of claim 3, in the measurement step, the brightness of the images of the deep grooves and the shallow grooves that the object to be inspected alternately has at regular intervals is measured. In the gain calculation step, a gain for correcting the brightness of the deep groove image is calculated by using the brightness obtained by averaging the brightness of the shallow groove images located on both sides of the deep groove. In the brightness correction stage, this gain is used to correct the brightness for each deep groove image. The brightness of the image of the shallow groove located on both sides of the deep groove has a linear relationship with the brightness of the image of the deep groove, and the brightness of the image of the deep groove and the brightness of the image of the shallow groove has a certain relationship. Moreover, the foreign material in the shallow groove is easily removed, and the possibility of having the foreign material is low. For this reason, the gain for correcting the brightness of the image for each deep groove is calculated by using the brightness obtained by averaging the brightness of the images of the shallow grooves on both sides of the deep groove, so that it is not affected by variations in illumination or the influence of foreign matter. The brightness of the deep groove can be estimated. Thereby, the recognition accuracy of the captured image in the deep groove can be improved, and a precise foreign object inspection can be performed.

  According to the fourth aspect of the invention, the conical optical system refracts the light that has passed through the shallow groove and the deep groove and enters the image input means. For this reason, it is possible to inspect the foreign matter in the deep part of the deep groove, which is difficult to measure by imaging from above the groove. By the way, by inserting a conical optical system in the center of the deep grooves and shallow grooves formed radially, and performing foreign object inspection, the brightness of the image of the deep grooves can be determined by the central axis of the conical optical system and the centers of the deep grooves and shallow grooves. It is affected by misalignment with the shaft. However, in the deep groove and the shallow groove that are radially arranged at a constant interval, the brightness of the image of the shallow groove located on both sides of the deep groove has a linear relationship with the brightness of the image of the deep groove. Therefore, the gain that corrects the brightness of the image for each deep groove is calculated by using the brightness that averages the brightness of the images of the adjacent shallow grooves. It is possible to estimate the brightness of the deep groove that has not received the light. Thereby, the recognition accuracy of the captured image in the deep groove can be improved, and a precise foreign object inspection can be performed.

  According to the fifth aspect of the present invention, the position adjusting mechanism abuts the tapered surface of the guide means having a tapered surface having an inner diameter extending from the image input means side toward the object to be inspected and the outer edge of the object to be inspected. The central axis of the optical system and the central axis of the shallow and deep grooves are adjusted coaxially. Thereby, attenuation of the light passing through the conical optical system is suppressed, and the recognition accuracy of the captured image can be improved.

(First embodiment)
A foreign substance inspection apparatus according to a first embodiment of the present invention is shown in FIG. The foreign matter inspection apparatus 20 of the present embodiment is a device that detects foreign matter in a deep groove that divides each segment of a commutator used in a fuel pump.
First, a commutator to be inspected and a fuel pump using the commutator will be described with reference to FIGS. The fuel pump 50 is an in-tank type Wesco pump mounted in a fuel tank of a vehicle or the like, for example. As shown in FIG. 9, the fuel pump 50 includes a pump unit 51, a motor unit 52 that rotationally drives the pump unit 51, and a discharge side cover 53.
The pump unit 51 includes an impeller 54, a pump cover 55 that accommodates the impeller 54, and a pump casing 56. The fuel sucked from the suction port 57 formed in the pump cover 55 is pressurized in the pump flow path 58 by the rotation of the impeller 54, passes between the permanent magnet 60 and the armature 61 of the motor unit 52, and is discharged from the discharge port. 72 is discharged.
The motor unit 52 constitutes a DC motor, and has a permanent magnet 60, an armature 61, and a commutator 80 on the side of the armature 61 opposite to the pump unit. The permanent magnet 60 attached to the inner peripheral wall of the housing 62 forms magnetic poles having different poles in the rotation direction.

  The armature 61 includes a shaft 64 as a rotating shaft member press-fitted into the central core 63. Six magnetic pole coil portions 65 are provided on the outer periphery of the central core 63 in the rotational direction, and are coupled to the central core 63. Each magnetic pole coil section 65 has a coil core 66, a bobbin 67, and a coil 68 formed by concentrating windings around the bobbin 67. The six magnetic pole coil portions 65 have the same configuration.

  The start and end of each coil 68 on the commutator 80 side are electrically connected to the terminal 69. The terminal 69 is engaged with and electrically connected to the connection terminal 85 of the commutator 80. On the other hand, the end of the coil 68 on the pump portion side is electrically connected to the terminal 70. Three terminals 70 that are adjacent to each other in the rotation direction are electrically connected by a terminal 71.

  The discharge-side cover 53 covers the commutator 80 side of the motor unit 52 and forms a discharge port 72 at a substantially central portion. A brush 73 is provided on the commutator 80 side of the discharge side cover 53. The brush 73 is urged by a spring 74 and is in contact with the commutator 80. A power receiving connector 75 is formed on the outer peripheral side shifted from the central portion of the discharge side cover 53. A terminal 76 provided in the power receiving connector 75 is electrically connected to the brush 73. A drive current is supplied from the terminal 76 to the commutator 80 via the brush 73.

As shown in FIGS. 10 and 11, the commutator 80 includes six segments 82 divided by six deep grooves 81 extending linearly in the radial direction. The segment 82 is made of carbon, for example. Each segment 82 has a shallow groove 83 as a waste oil groove in the middle in the rotation direction. The deep grooves 81 and the shallow grooves 83 are radially formed at equal intervals.
The armature side of the segment 82 is in contact with and electrically connected to the intermediate terminal 84. The intermediate terminal 84 is made of, for example, brass. The intermediate terminal 84 is coupled to the connection terminal 85 by the protrusion 87 and is electrically connected. The connection terminal 85 is formed in an arc shape when viewed from the axial direction, and electrically connects the intermediate terminals 84 opposed in the radial direction. Thereby, the segments 82 provided facing each other in the radial direction have the same potential.
When the drive current is supplied from the terminal 76 of the power receiving connector 75 to the segment 82 via the brush 73, the drive current is supplied from the segment 82 to the coil 68 via the intermediate terminal 84, the connection terminal 85, and the terminal 69. The armature 61 rotates. Due to the rotation of the armature 61, each segment 82 sequentially contacts the brush 73.

In the manufacturing process of the commutator 80, first, the base material of the segment before dividing into six pieces, the base material of the intermediate terminal, and the connection terminal 85 are assembled and insert-molded with the insulating resin 86. Next, the deep groove 81 is formed by cutting the base material of the segment and the base material of the intermediate terminal with a disk-shaped cutter (not shown). By forming the deep groove 81, the segment 82 and the intermediate terminal 84 are divided into six and are electrically insulated. The thickness of the disk-type cutter is, for example, 0.35 mm ± 0.01 mm. As shown in FIG. 12, the deep groove 81 is formed with a width a of, for example, 0.35 mm ± 0.05 mm and a depth b of, for example, 4 mm. A cut section 841 of the intermediate terminal 84 is exposed in the deep portion of the deep groove 81. Chips and the like resulting from cutting the intermediate terminal 81 become foreign matter in the deep groove.
Furthermore, the shallow groove | channel 83 used as a waste oil groove | channel is formed by cut | disconnecting the intermediate | middle of the rotation direction of each segment 82 with a disk type cutter. The shallow groove 83 is formed with a width of, for example, 0.35 mm ± 0.05 mm and a depth of, for example, 2.5 mm.

Next, the configuration of the foreign matter inspection apparatus 20 will be described. As shown in FIG. 1, the foreign substance inspection apparatus 20 includes an image processing unit 21, an image input unit 22, a conical optical system 26, a light source 29, a position adjusting mechanism, and the like.
The image processing means 21 includes a computer having a CPU, a memory such as a RAM and a ROM, and peripheral devices thereof, and is connected to the image input means 22. The image processing means 21 controls the image input means 22 by loading a foreign substance inspection program stored in a ROM or the like into the CPU, and performs image processing on the image data transmitted from the image input means 22 to inspect the object to be inspected. The presence or absence of foreign matter in the deep groove of the commutator 80 is measured.

The image input means 22 is a camera configured to include a lens for capturing an image and an image sensor such as a CCD, and is attached to the mounting arm 24 of the column 23. The image input means 22 photoelectrically converts the light incident from the conical optical system and transmits it to the image processing means 21.
The conical optical system 26 is configured by a prism in which a cylindrical base 261 and a conical tip 262 are integrally formed. The conical optical system 26 is attached to a support arm 25 provided in the middle of the column 23. An inclination guide 27 is provided at an intermediate position in the axial direction of the conical optical system 26 to adjust the inclination of the axis of the conical optical system 26.

The light source 29 is annular illumination that irradiates light toward the groove of the object to be inspected, and is provided so as to surround the entire circumference of the commutator 80 from the radially outward direction of the commutator 80.
The position adjustment mechanism includes a guide 31, a lifting platform 34, and a pallet 32.
The guide 31 as the guide means has a conical tapered surface 36 whose inner diameter increases toward the commutator 80 as the object to be inspected, and is fixed to the support arm 25. The masking 28 is provided on the inner peripheral side of the tapered surface 36 of the guide 31 and is movable in the vertical direction.
The elevator 34 has a pallet 32 at the top and is installed on a conveyor 35. A commutator 80 is placed on the pallet 32. The lifting platform 34 includes a cylinder 33 therein, and can move the pallet in the vertical direction. The guide 31, the lifting platform 34, and the pallet 32 constitute a position adjusting mechanism in the scope of claims.

Next, the operation of the foreign matter inspection apparatus 20 will be described.
When the elevator 34 with the commutator 80 placed on the pallet 32 is moved directly below the conical optical system 26 by the conveyor 35, the movement of the elevator 34 is stopped by a stopper (not shown). Next, the cylinder 33 inside the lifting platform 34 extends in the upward direction, and the pallet 32 on which the commutator 80 is placed moves to the conical optical system 26 side. At this time, it is assumed that the center axis t of the commutator 80 and the center axis s of the conical optical system 26 are slightly shifted as shown in FIG. As the cylinder 33 extends in the top direction, the outer edge 89 of the commutator 80 and the tapered surface 36 of the guide 31 come into contact with each other. Further, the cylinder 33 extends slightly in the upward direction, and as shown in FIG. 2B, the commutator 80 is guided by the tapered surface 36, and the central axis t of the commutator 80 and the central axis s of the conical optical system 26 are It is adjusted to be substantially coaxial. At this time, the center axis t of the commutator 80 and the center axis s of the conical optical system 26 may be slightly shifted due to the tolerance between the outer edge 89 of the commutator 80 and the deep grooves 81 and the shallow grooves 83.
Next, the masking 28 moves toward the commutator 80, and comes into contact with the upper surface in the axial direction of the segment 82 and the insulating resin 86 of the commutator 80. Thereby, in the deep groove 81 and the shallow groove 83 of the commutator 80, the penetration of light from the axial direction is blocked.

  The light emitted from the light source 29 passes through the deep groove 81 and the shallow groove 83 from the radially outward direction of the commutator 80 and enters the conical optical system 26. As shown in FIG. 3, among the light passing through the deep groove 81, the light 40 passing through the axial surface side of the commutator 80 is refracted by the conical surface of the distal end portion 262 of the conical optical system 26, and the outer diameter of the base portion 261 is reflected. Then, the light enters the image input means 22. Of the light that passes through the deep groove 81, the light 41 that passes through the deep side in the axial direction of the commutator 80 is refracted by the conical surface of the tip 262 and enters the image input means 22 through the inside of the base 261.

  When the central axis s of the conical optical system 26 and the central axes t of the deep groove 81 and the shallow groove 83 of the commutator 80 are the same, the light 40 passing through the surface side of the deep groove 81 is incident on the tip 262 of the conical optical system 26. The optical path to be performed is shown in FIG. When the conical optical system 26 is viewed from the axial image input means side, both the transmitted light 43 directly transmitted through the deep groove 81 and the reflected light 44 reflected by the side wall 88 forming the deep groove 81 are both conical optical system 26. And refracted at the conical surface opposite to the deep groove 81 and enters the image input means 22. At this time, the transmitted light 43 is bright and the reflected light 44 is darker than the transmitted light 43. Therefore, when the center axis of the commutator 80 and the conical optical system 26 is misaligned, the range of the image by the transmitted light 43 is reduced depending on the direction and distance of the misalignment and the positional relationship between the deep grooves 81 and the reflected light. The range of the image by 44 is enlarged.

  The image input means 22 photoelectrically converts the light incident from the conical optical system 26 and transmits it to the image processing means 21. Image data input to the image processing means 21 is shown in FIG. In FIG. 5, only transmitted light is shown and the reflected light is not shown to simplify the description. In FIG. 5, it is assumed that the inside of each triangle is bright and the entire circle is dark. Triangular B image 2, D image 4, F image 6, H image 8, J image 10 and L image 12 show images of each deep groove 81, respectively, and triangular A image 1, C image 3, E image 5, The G image 7, the I image 9, and the K image 11 show images of the respective shallow grooves 83, respectively. In each triangle image, the vertex on the center side of the circle indicates the deep side of the deep groove 81 or the shallow groove 83, and the base of each triangle image located in the radially outward direction of the circle is the surface side of the deep groove 81 or the shallow groove 83. Is shown. The area of the triangle showing the A image 1 and the G image 7 of the shallow groove 83 is smaller than the area of the triangle showing the images 3, 5, 9, and 11 of the other shallow grooves 83. This indicates that the A image 1 and the G image 7 of the shallow groove 83 are darker than the images 3, 5, 9, and 11 of the other shallow grooves 83. Further, the area of the triangle indicating the B image 2 and the H image 8 of the deep groove 81 is smaller than the area of the triangle indicating the images 4, 6, 10, and 12 of the other deep grooves 81. This indicates that the B image 2 and the H image 8 of the deep groove 81 are darker than the images 4, 6, 10, and 12 of the other deep grooves 81.

The image processing means 21 performs image processing on the captured image data in order to measure the presence or absence of foreign matter in the deep groove 81. The procedure of image processing performed by the image processing means 21 is shown in the flowchart of FIG.
First, the image processing means 21 performs a scaling process for adjusting the entire image to appropriate brightness for the captured image data (S101). Next, the center positions of the images of the deep grooves 81 and the shallow grooves 83 captured radially are detected (S102). Then, using the center position detected in S102 as a reference, the radial image is converted into a strip image (S103).
An image of the shallow groove 83 and the deep groove 81 in this state is shown in FIG. As in FIG. 5, it is assumed that the inner part of the triangle is bright and the background is dark. The A image 1 and the G image 7 of the shallow groove 83 are darker than the images 3, 5, 9, and 11 of the other shallow grooves 83. Further, the B image 2 and the H image 8 of the deep groove 81 are darker than the images 4, 6, 10, 12 of the other deep grooves 81.

The image processing means 21 corrects the brightness of the image of the deep groove 81 by a brightness correction method described later (S104). FIG. 8B shows an image of the deep groove 81 and the shallow groove 83 after the luminance correction is performed. The area of the triangle indicating the B image 2 and the H image 8 of the deep groove 81 is the same size as the other deep groove 81 images, and the B image 2 and the H image 8 of the deep groove 81 are affected by light attenuation and the like. It shows that it approximated to the brightness that was not received. The foreign substance region 13 is shown in the H image 8 of the deep groove 81.
Next, the image processing means 21 creates an inspection area for foreign object inspection, and performs binarization processing with a predetermined threshold (S105). In the present embodiment, the inspection area is set to 1.66 mm from the deep part of the deep groove 81, for example. In the commutator 80, an inspection region of 0.4 mm from the deep portion is an effective region for preventing conduction between adjacent segments.

  The image processing means 21 performs a morphology such as a closing operation or an opening operation to remove noise in the foreign substance region (S106). Next, a convex hull region is created and used as an image representing the original shape of the deep groove (S107). Then, a difference between the image obtained in S107 and the image obtained in S106 is detected (S108). The presence / absence of a foreign object is measured from the difference thus obtained (S109).

Next, the procedure of the luminance correction method in S104 will be described based on the flowchart of FIG.
First, the image processing means 21 measures the brightness of the images 1, 3, 5, 7, 9, 11 of each shallow groove 83 and the brightness of the images 2, 4, 6, 8, 10, 12 of each deep groove 81, respectively. (S201). Here, due to the misalignment between the center axis s of the conical optical system and the center axis t of the deep groove 81 and the shallow groove 83, the inclination of the conical optical system, the composition of the conical optical system, and the foreign matter included in the deep groove 81, etc. The brightness of the images of the deep groove 81 and the shallow groove 83 is not uniform. However, the brightness of the image of the deep groove 81 and the brightness of the image of the shallow groove 83 have a predetermined relationship, and the brightness of the image of each deep groove 81 has a linear relationship with the brightness of the adjacent shallow grooves 83. . Therefore, the image processing means 21 calculates a gain for correcting the luminance of the image of each deep groove 81 using the luminance obtained by averaging the luminance of the adjacent shallow grooves 83 (S202). For example, a gain that corrects the luminance of the B image 2 in the deep groove 81 is calculated by targeting the luminance obtained by averaging the luminances of the A image 1 and the C image 3 in the shallow groove 83. Further, the gain for correcting the brightness of the image of the other deep groove 81 is calculated with the brightness obtained by averaging the brightness of the images of the shallow grooves 83 located on both sides of each deep groove 81 as a target.
Then, the image processing unit 21 corrects the luminance for each image of the deep groove 81 using the gain calculated in S202 (S203). In the present embodiment, it is possible to detect a foreign object having a diameter of 0.2 mm or less.

In the present embodiment, the gain for correcting the brightness of the image for each deep groove 81 is calculated using the brightness obtained by averaging the brightness of the images of the shallow grooves 83 on both sides of the deep groove 81. In linear deep grooves 81 and shallow grooves 83 that are alternately arranged at regular intervals, the luminance of the image of the deep groove 81 and the luminance of the image of the shallow groove 83 located on both sides of the deep groove 81 have a linear relationship. . Therefore, the center axis s of the conical optical system and the center axis t of the deep groove 81 and the shallow groove 83 are decentered, the inclination of the conical optical system, the composition of the conical optical system, and the influence of illumination or foreign matter. The brightness of the deep groove 81 that is not present can be estimated. For this reason, it can suppress that the foreign material area | region of the image of the deep groove 81 is assimilated with the image of the background or the deep groove 81 in the case of a binarization process. As a result, the recognition accuracy of the captured image of the deep groove 81 can be improved, and a precise foreign object inspection can be performed.
Furthermore, in this embodiment, the tapered surface 36 of the guide 31 and the outer edge 89 of the commutator 80 are brought into contact with each other, and the central axis s of the conical optical system 26 and the central axes t of the shallow groove 83 and the deep groove 81 are adjusted coaxially. is doing. Thereby, the attenuation of light passing through the conical optical system 26 is suppressed, and the recognition accuracy of the captured image can be improved.

(Second Embodiment)
A foreign substance inspection apparatus according to a second embodiment of the present invention is shown in FIG. Components substantially the same as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted. In 2nd Embodiment, the foreign material in the groove | channel 91 of the to-be-inspected object 90 formed in the rectangular parallelepiped is detected. The inspected object 90 has linear M grooves to Q grooves 91 at equal intervals. The depth, width, and length of each groove 91 are the same size.
Light emitted from a light source (not shown) passes through the grooves 91 from the direction of the arrow 45 and enters the image input means 22. The image input means 22 photoelectrically converts the light incident from the conical optical system 26 and transmits it to the image processing means 21. Note that masking (not shown) may be attached to the upper surface 92 of the object to be inspected to block the incidence of light from above.

The image repair means 21 performs image processing according to the procedures of S101 to S109 and S201 to S203 described in the first embodiment, and measures the presence or absence of foreign matter in the groove 91. In the present embodiment, the procedure of S201 to S203 will be described in more detail with reference to FIGS.
The image processing means 21 measures the brightness of the image of each groove 91 (S201). FIG. 14 shows a waveform input based on the brightness of the image of the M groove to Q groove 91 measured by the image processing means 21. In FIG. 14, the coordinate on the horizontal axis corresponds to the coordinate on the horizontal axis of the image converted into a band shape in S103. The density value on the vertical axis corresponds to the average luminance of pixels that are continuous in the vertical axis direction in this strip-shaped image.
The image processing means 21 calculates an average value for each waveform corresponding to the images of the M groove to the Q groove 91. The average value of this waveform is shown in FIG. Next, the image processing means 21 corrects the luminance of the image of the M-groove to Q-groove 91, and the average value of the waveform corresponding to the image of the M-groove to Q-groove 91 is set to the middle of 256 gradations. An average value adjustment process is performed according to the value. The waveform after this processing is shown in FIG.

The image processing means 21 calculates a gain for correcting the luminance of the image of the predetermined groove using the luminance obtained by averaging the luminances of the images of the grooves located on both sides of the predetermined groove (S202). For example, the gain for correcting the N-groove image is calculated using the luminance obtained by averaging the luminance of the image of the M-groove 91 and the image of the O-groove 91. Similarly, a gain for correcting the image of the O-groove is calculated using the luminance obtained by averaging the luminance of the image of the N-groove 91 and the image of the P-groove 91. A gain that corrects the image of the P groove is calculated using the luminance obtained by averaging the luminance of the image of the O groove 91 and the image of the Q groove 91.
The image processing means 21 uses the gain calculated in S202 to correct the brightness for each image of the M-groove to Q-groove 91 (S203). FIG. 16 shows a state in which the waveform corresponding to the image of the M groove to the Q groove 91 is stretched in the vertical axis direction by this correction. The waveform corresponding to the images of the M-groove to Q-groove 91 is relatively stretched from the waveform indicated by the solid line to the waveform indicated by the dotted line with reference to a predetermined value. FIG. 17 shows a state where the brightness correction is completed in this way. The waveform corresponding to the image of the M-groove to Q-groove 91 has a distribution of density values over a wide range.

In the present embodiment, the gain for correcting the luminance of the image for each predetermined groove is calculated using the luminance obtained by averaging the luminances of the images of the grooves adjacent to the predetermined groove. In linear grooves arranged at regular intervals, the brightness of an image of a predetermined groove and the brightness of the images of grooves located on both sides of the predetermined groove have a linear relationship. For this reason, the brightness | luminance of the predetermined groove | channel which is not influenced by the dispersion | variation in illumination etc. can be estimated.
Further, the image processing unit 21 corrects the luminance for each image of the M-groove to Q-groove 91 after performing an average value adjusting process for matching the average value of the images of the M-groove to Q-groove 91 with a predetermined value. For this reason, the luminance of the M-groove to Q-groove 91 images is relatively corrected with reference to a predetermined value. As a result, the threshold margin in the binarization process can be increased, and the foreign matter area can be prevented from being assimilated into the background or groove image. As a result, it is possible to improve the recognition accuracy of the captured image of the groove and perform a precise foreign object inspection.

(Other embodiments)
In the first embodiment, the image processing method for detecting the foreign matter in the deep groove 81 of the commutator 80 having the deep groove 81 and the shallow groove 83 has been described. On the other hand, the present invention may be applied to an image processing method for detecting foreign matter in the deep groove of the commutator 80 having only the deep groove 81.
In the second embodiment, the image processing method for detecting the foreign matter in the groove 91 of the inspected object 90 formed in a rectangular parallelepiped has been described. On the other hand, the present invention may be applied to an inspected object having various shapes as long as the inspected object has three or more linear grooves at regular intervals, and has deep grooves and shallow grooves alternately. You may apply this invention to a to-be-inspected object.
As described above, the present invention is not limited to the above-described embodiment, and can be implemented in various forms without departing from the gist thereof.

Sectional drawing of the foreign material inspection apparatus by 1st Embodiment of this invention. (A), (B) The figure which shows the action | operation of the foreign material inspection apparatus by 1st Embodiment of this invention. The figure which shows the optical path by the conical type optical system by 1st Embodiment of this invention. The figure which shows the optical path by the conical type optical system by 1st Embodiment of this invention. The schematic diagram which shows the image of the shallow groove | channel and deep groove | channel by 1st Embodiment of this invention. 3 is a flowchart showing an image processing method according to the first embodiment of the present invention. 3 is a flowchart showing an image processing method according to the first embodiment of the present invention. (A), (B) The schematic diagram which shows the image of the shallow groove | channel and deep groove | channel by 1st Embodiment of this invention. Sectional drawing of the fuel pump with which the commutator used as the test object of the foreign material inspection apparatus by 1st Embodiment of this invention is used. The perspective view of the commutator used as the test object of the foreign material inspection apparatus by 1st Embodiment of this invention. Sectional drawing of the commutator used as the test object of the foreign material inspection apparatus by 1st Embodiment of this invention. The principal part sectional drawing of the commutator used as the test object of the foreign material inspection apparatus by 1st Embodiment of this invention. The schematic diagram of the foreign material inspection apparatus by 2nd Embodiment of this invention. The graph which shows the luminance distribution of the image by 2nd Embodiment of this invention. The graph which shows the luminance distribution of the image by 2nd Embodiment of this invention. The graph which shows the luminance distribution of the image by 2nd Embodiment of this invention. The graph which shows the luminance distribution of the image by 2nd Embodiment of this invention.

Explanation of symbols

20: foreign matter inspection device, 21: image processing means, 22: image input means, 23: support, 24: mounting arm, 25: support arm, 26: conical optical system, 27: tilt guide, 28: masking, 29: Light source, 31: guide, 32: pallet, 33: cylinder, 34: lifting platform, 35: conveyor, 80: commutator (inspection object)

Claims (5)

An image processing method for measuring the presence or absence of foreign matter in the groove based on image data obtained by photoelectrically converting light passing through the groove of an object to be inspected having three or more linear grooves at regular intervals. And
A measurement stage for measuring the brightness of the image for each groove;
A gain calculating step of calculating a gain for correcting the brightness of the image of the predetermined groove using a brightness obtained by averaging the brightness of the images of the grooves located on both sides of the predetermined groove;
Using the gain, a gain correction step of correcting the brightness for each image of the predetermined groove,
A computer-processed image processing method. The image processing method by a computer according to claim 1.
An image processing method by a computer, further comprising an average value adjustment step of correcting the luminance of the image for each groove and adjusting an average value of the luminance of the image for each groove to a predetermined value. The image processing method by a computer according to claim 1 or 2,
The object to be inspected has deep grooves and shallow grooves formed in a straight line alternately at regular intervals,
In the measurement stage, the brightness of the image of the deep groove and the shallow groove is measured,
In the gain calculation step, a gain for correcting the brightness of the image of the deep groove is calculated using a brightness obtained by averaging the brightness of the image of the shallow groove located on both sides of the deep groove,
In the brightness correction stage, the gain is used to correct the brightness for each image of the deep groove, and the image processing method by a computer is characterized in that: A light source for irradiating light into the groove of the object to be inspected, which has the deep grooves and the shallow grooves formed linearly at a constant interval; and
A conical optical system that is inserted in the center of the deep groove and the shallow groove and refracts light that has passed through the shallow groove and the deep groove;
Image input means for photoelectrically converting light incident from the conical optical system;
Using the image processing method according to claim 3, image processing means for measuring the presence or absence of foreign matter in the deep groove based on image data transmitted from the image input means;
A foreign matter inspection apparatus comprising: The foreign matter inspection apparatus according to claim 4,
The tapered surface of the guide means having a tapered surface having an inner diameter extending from the image input means side toward the object to be inspected and the outer edge of the object to be inspected, the central axis of the conical optical system, the shallow groove, A foreign matter inspection apparatus further comprising a position adjustment mechanism that adjusts the central axis of the deep groove coaxially.

Images