JPH09229645A - Gear chamfer face measuring device and chamfer face irradiating method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To establish an accurate judgement of the acceptability of gear chamfering workmanship worked to the end faces of each tooth of gear, which is made through an image measurement of the symmetry amount to the two chamfer faces. SOLUTION: A one-side image to represent a chamfer face on one surface is extracted among picked-up images (S1), and a regression equation of a presumed ridge line presumed along approx. the ordinate axis is determined on the basis of the obtained one-side image (S2). The position of the addendum circle is located on the basis of the picked-up image determined in the direction of the regressive straight line which is defined by the regression equation (S3), and a reference line is set perpendicular to the regressive straight line as positioned inward for a prescribed distance from the position of the addendum circle (S4). The picture element value distribution of the picked-up image is determined in the direction of the set reference line, and the picture element value varying point to represent the position of ridge line and two picture element value varying points to represent the contour positions of the left and right tooth flanks are decided from the obtained picture element value distribution, and the symmetry amount is calculated from these picture element value varying points to serve for judgement of the acceptability of the chamfering workmanship (S4-S7).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a gear that moves in the tooth trace direction to change gears, such as the inner teeth of a sleeve and the outer teeth of a synchronizer ring in a transmission. The present invention relates to a gear chamfered surface measuring device and a chamfered surface illuminating method for measuring a distribution amount of a chamfered surface formed with a taper by a new method and determining pass / fail.

[0002]

2. Description of the Related Art In a manual transmission gear shift, when a clutch is disengaged and a sleeve is moved by a shift fork, a synchronizer key fitted with a sleeve and a protrusion moves together with the sleeve, and a synchronizer ring is attached to the end face of the key to connect a synchronizer ring to a cone of a third gear. Press on the part. When the sleeve progresses and the sleeve and the synchronizer key are disengaged from each other, the internal teeth (splines) of the sleeve and the external teeth (splines) of the synchronizer ring that are out of phase with each other in the circumferential direction (rotational direction) The synchronizer ring blocks the movement of the sleeve due to the interference between the end faces, and the sleeve directly presses the synchronizer ring, and the synchronizer ring strongly presses against the cone portion of the third gear (synchronization action). After this, when the third gear becomes the same as the rotation of the sleeve and the synchronizer ring, the synchronizer ring becomes free and does not block the movement of the sleeve, and the sleeve passes through the synchronizer ring to complete the mesh gear shift with the spline of the third gear.

Therefore, as shown in FIG. 15, chamfered surfaces are formed at both end portions in the tooth height direction on the end faces of the inner teeth of the sleeve and the outer teeth of the synchronizer ring which are generally used in a transmission, as shown in FIG. , The sleeve moves smoothly to the third gear side with respect to the synchronizer ring. Here, in order to make the movement of the sleeve smooth, it is desirable that the double-sided chamfered surface be ground by substantially the same amount. That is, as shown in FIG. 16, the ridge line where the chamfered surfaces intersect with each other matches that the center A in the tooth trace direction of each tooth is that the synchronizer ring crimped to the cone portion of the third gear and the sleeve are engaged with each other. Will be secured and lead to early synchronization.

Here, as a parameter showing the degree of coincidence of the ridgeline of the double-sided surface with the center A of each tooth, FIG.
As shown in (A), for example, the ridge line position C on the pitch circle P
And the distribution amount a / b (or b / a) represented by the ratio of the distances a and b when the left and right tooth flank positions B1 and B2 are viewed from the front of the double-sided chamfer. The measuring device for the distribution amount of the chamfered surface irradiates the chamfered side end surface of the gear to be inspected, which has been conventionally installed on the measurement jig, with a ring-shaped illumination means such as a surfline fluorescent lamp with a small reflection, and the chamfered surface is Evenly illuminated to obtain a captured image, the pixel value distribution on the pitch circle P (see FIG. 17).
(See (B)), the ridge line position C and the left and right tooth flank positions B1,
B2 is determined and the distribution amount is measured.

[0005]

In the conventional gear chamfering surface measuring device, when the focus of the image pickup means is adjusted to the ridge line position C, the pixel value distribution of the left and right tooth flank positions B1 and B2 becomes unclear, and the accurate distribution is performed. Unable to calculate quantity. This is the same when focusing on the left and right tooth flank positions B1 and B2.

Further, in the conventional gear chamfering chamfering device, the position of the pitch circle P is set to a fixed position which is parallel-shifted with respect to the visual field boundary or the tip circle position, so that the gear to be inspected is on the jig. When set tilted to, the distances a and b
Is not obtained, and the accuracy of distribution amount measurement deteriorates. The present invention
By extracting only one side chamfered surface from the captured image and determining the left and right tooth flank positions and the ridge line position based on the extracted one sided image, a chamfered chamfering device for gears that improves the distribution amount measurement accuracy is provided. Providing is a problem to be solved.

Further, according to the present invention, the chamfered chamfered surface of the gear is measured by illuminating the chamfered surface so that there is a difference in brightness, thereby obtaining a clear picked-up image and further enhancing the accuracy of the distribution amount measurement. It is an issue to be solved to provide a lighting method for the purpose.

[0008]

The gist of the invention of claim 1 which has solved the above-mentioned problems is (a) an image pickup means for picking up a double-sided surface to be machined on the end face of each tooth of a gear; A single-sided chamfer extraction means for extracting a single-sided image representing a single-sided chamfer from the imaged image obtained by the image-capturing means; and (c) substantially orthogonal to the single-axis based on a plurality of parallel uniaxial pixel value distributions of the single-sided image. Regression equation setting means for obtaining a regression equation of an estimated ridge line estimated in a direction, (d) tip circle position determining means for obtaining a tip circle position based on a regression line defined by the above regression equation, and (e) Reference line setting means for setting a reference line orthogonal to the regression line inward by a predetermined distance from the tip circle position determined by the tip circle position determining means,
(F) A pixel value distribution of the captured image is obtained in the direction of the reference line set by the reference line setting means, and a pixel value change point indicating a ridge line position and two pixels indicating both contour positions are obtained from the pixel value distribution. Distribution amount determination point calculation means for determining a value change point, and (g) calculation means for calculating a distribution amount from each of the pixel value change points determined by the distribution amount determination point calculation means,
(H) A determination means for determining whether the gear is good or bad is provided by processing the chamfered surface based on the result of the calculation means.

The chamfered chamfer measuring device for gears according to the first aspect of the present invention calculates the distribution amount by the procedure shown in FIG. In FIG. 1, step S1 is a process performed by the one-side chamfered surface extraction unit,
S2 is a process performed by the regression equation setting unit, S3 is a process performed by the tip circle position determining unit, S4 is a process performed by the reference line setting unit,
S5 is a process performed by the distribution amount determination point calculation unit, S6 is a process performed by the calculation unit, and S7 is a process performed by the determination unit.

A picked-up image obtained by the pick-up means is a multi-gradation gray-scale image, and the one-side chamfered surface extraction means discriminates each gradation representing the double-sided chamfer by using a specific level of the picked-up image as a threshold value. , A single-sided image is extracted (S1). The regression equation setting means reads pixel values of the one-sided image along a plurality of horizontal axis directions on the screen to obtain a plurality of horizontal axis direction pixel value distributions.
Then, a linear regression equation is defined based on the coordinate values of a plurality of points at which the pixel values of the plurality of pixel value distributions in the horizontal axis direction change abruptly. This regression equation represents the ridge line whose existence is estimated to be substantially in the vertical axis direction as a linear function on the screen (S2).

The tip circle position determining means obtains the pixel value distribution of the captured image along the direction of the regression line determined by the above regression equation, for example, on the regression line, and the point where the pixel value suddenly changes is defined as the tip circle position. Yes (S3). When the tip circle position is determined, the image of the double-sided surface always exists inward of the tip circle position. Therefore, the reference line setting means is located at a position inward from the tip circle position by a predetermined distance. If a regression equation of a reference line orthogonal to the regression line is set, the reference line can be obtained as a straight line intersecting with the actual ridgeline (S4).

Therefore, the distribution amount determination point calculation means reads the image picked up by the image pickup means along the reference line, and from the pixel value distribution obtained by this, the pixel value change points and the left and right tooth flank contour positions which represent the ridge line position are obtained. Two pixel value change points to be represented can be determined (S5). When the coordinates of the ridge line position and the left and right tooth flank contour positions are obtained, the distribution amount RATEMAT is CPCL for the ridge line position on the reference line and CP for the left and right tooth flank contour positions.
As L and CPR,

[0013]

[Equation 1] However, it is obtained by calculating CPL <CPCL <CPR (S6).

The obtained distribution amount is judged by the judging means (S7). As described above, according to the first aspect of the invention, the regression line in which the ridge line is accurately estimated is obtained based on the single-sided image showing the single-side chamfered surface, and the ridge line position and the left and right tooth flank contour positions are determined from the regression line. Since the reference line is determined, the distribution amount is calculated remarkably correctly, and the accuracy of pass / fail of the distribution amount determination is improved.

Here, the position which is inward from the tip circle position by a predetermined distance is preferably a pitch circle position. The gist of the invention of claim 2 is that the one-sided chamfer extraction means is a binary image in which the maximum area area of the same pixel value of the captured image excluding the background portion obtained by the imaging means and the other area are divided into shades. It is characterized in that it is maximum area extraction means for extracting the one-sided image.

The gist of the invention of claim 3 is to use a chamfered chamfering device for gears including the maximum area extraction means of claim 2 as the one-side chamfered chamfer extraction means, and to illuminate the chamfered chamfers so that there is a difference in brightness. To do. According to the third aspect of the present invention, the gear to be inspected is illuminated so as to cause a difference in brightness of the double-sided surface, so that the captured image obtained by the image capturing means has a contrast of light and shade representing the double-sided surface with the ridge line as a boundary. Becomes extremely clear. Therefore, as in the maximum area extraction means of claim 2, in the operation of dividing the maximum area area of the same pixel value of the captured image excluding the background portion and the other area into shades to form a binary image as a single-sided image, The threshold value can be set to a value that can reliably discriminate the one-side chamfered surface, and the one-sided image is a more faithful representation of the actual one-side chamfered surface. Sex is obtained.

[0017]

As described above, according to the present invention, the ridge line is estimated based on the single-sided image showing the single-side chamfered surface and the regression line thereof is obtained. Therefore, the regression line depends on the inclination of the gear to be inspected. Since it can be obtained accurately, the pitch circle (reference line) can be surely parallel to the addendum circle, and the ridge line position and the left and right tooth flank contour positions determined from the captured image can be determined with a small error, so that the sorting can be performed. It is possible to improve the accuracy of pass / fail of the amount determination.

In particular, by adopting the illumination method according to the third aspect, the one-sided image can be more accurately extracted, and the accuracy of the distribution amount pass / fail judgment is further improved.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION The chamfered chamfering surface measuring device and chamfered chamfered illumination method of the present invention will be described in detail below. FIG.
Shows an overall configuration of a chamfered surface measuring device for a gear that employs the chamfered surface illumination method of the present invention. As shown in FIG. 2, the gear piece 1 to be inspected is housed in a sealed box 2 that shields external light and measured. A jig 3 for installing the gear piece 1 to be inspected, an illuminating means 4 for illuminating the gear piece 1 to be inspected on the jig 3, and an image of the gear piece 1 to be inspected illuminated by the illuminating means 4 are imaged in the closed box 2. And an image pickup means 5 for performing the operation.

In detail, the gear piece 1 to be inspected is on a jig 3 installed on the bottom plate of the hermetically sealed box 2 so that the end faces of the chamfered teeth face the image pickup means 5 consisting of the CCD camera 6 and the close-up lens 7. Fixed to. Here, the illuminating means 4 is adapted to illuminate the respective chamfered surfaces A and B with the light guides 4a and 4b formed of respective optical fibers with the illumination light emitted from the external halogen light source 9. A condenser lens 8 is provided at the tip of the light guide 4a that illuminates one chamfered surface A (left side in the drawing), and the chamfered surface A is the chamfered surface B.
It is illuminated more brightly.

An image processing system for an image pickup signal output from the CCD camera 6 includes an image processing apparatus 10 having an image reading program, an image display program, an image conversion program, an image feature calculation program and a distribution amount pass / fail judgment program. A host computer 11 that comprehensively manages the results determined by the distribution amount pass / fail determination program of the image processing apparatus 10, and a communication control apparatus 12 that performs control program development and operation management of the image processing apparatus 10 are mainly configured. A TV monitor 13, a key input board 14, and a video printer 15 are connected to the image processing apparatus 10. The communication control device 12 uses, for example, a personal computer 9801T manufactured by NEC, and activates the communication software WTERM.

The image processing apparatus 10 issues a turning command for turning the jig 3 to the servo driver 16 and controls the servo motor 17 for turning the jig 3 via the servo driver 16 to inspect the gear piece to be inspected. Rotate and position 1 per tooth. Next, the distribution amount acceptance / rejection determination operation performed by the image processing system will be described based on FIG. 3 which further details steps S1 to S7 in FIG.

S1 [Extraction of single-sided image] First, the gear piece 1 to be inspected is set on the jig 3 with the chamfered end face side facing upward, and the light guides 4a and 4b of the illumination means 4 are double-sided for each tooth. The irradiation angle and the like of the illumination means 4 are adjusted so that the surfaces A and B are illuminated with their respective illumination capabilities. As a result, the chamfered surface A is illuminated with higher brightness than the chamfered surface B.

Therefore, C read by the image processing apparatus 10 in step S11 [reading of captured image] in FIG.
As shown in FIG. 4A, the image picked up by the CD camera 6 is an image in which the chamfered surface A is the brightest, and the chamfered surface B and the background portion C have lower brightness in this order. The read captured image is subjected to the image processing of S12 [smoothing]. This smoothing is performed by i (i = 1, 2, 1) adjacent to the pixel Pn to be processed.
... 8) The product sum average of the pixel values Pi (X, Y) for each pixel is set as the nth pixel value Qn (X, Y).

Specifically, This reduces random noise due to uneven lighting.

Next, the smoothed picked-up image is S13 [2
Convert to binarized image by binarization conversion]. Specifically, the captured image shown in FIG. 4A is read at a position of the vertical axis Y = 150 (a position that surely passes through the double-sided chamfered surfaces A and B).
As shown in FIG. 5, a pixel value distribution in which the high-luminance side is high density is obtained, the maximum value TH “255” is obtained, and the maximum value TH
A binary image consisting of a dark part and a light part with a threshold value (see FIG. 5).
See).

In step S13, morphological reduction / expansion processing is applied to the binary image in two steps to remove small-scale noise and medium-scale noise due to surface irregularities. Further FIG.
In the binary image of, the dark area with the largest area is extracted,
The binary image is as shown in FIG. Specifically, in the image memory, if "0" is a dark portion and "1" is a light portion,
All pixel cells are set to "1", leaving only the pixel cell group having the maximum area in which "0" s are continuous. Thus, chamfered surface A
A single-sided image representing is extracted.

S2 (setting regression equation of estimated ridge line) Next, the ridge line formed by the chamfered surface is estimated. As the estimated ridge line, the single-sided image of FIG. 6 has at least two values of the vertical axis Y, here, for example, Y = 100, 200, 300, 40 in FIG.
The value PX (i) of the horizontal axis X at which the pixel value distribution read at the position of 0 changes from “0” to “1” is calculated, and the coordinate value [PY
(100), PX (1)], [PY (200), PX
(2)], [PY (300), PX (2)], [PY
(400), PX (3)] to set the regression equation.

That is, XM is the average value of the ridge line positions, and YM is Y.
When defined as the average value of axis coordinate values, the correlation coefficient GMR and the slope of the regression equation (regression coefficient) are calculated by the following statistical method. ^{SY = (100-YM) 2} + (200-YM) 2 + (300-YM) 2 = 50000 SXY = (100-YM) × (PX (0) -XM) + (200-YM) × (PX ( 1) -XM) + (300-YM) * (PX (2) -XM) + (400-YM) * (PX (3) -XM) =-150 * (PX (0) -XM) -50 * (PX (1) -XM) + 50 * (PX (2) -XM) + 150 * (PX (3) -XM) GMR = SXY / (SX * SY) < ^1/2 > BB = SXY / SX.

Here, when GMR ² > 0.9, the regression equation is as follows. Y = YM + BB × (X−XM) ……………………………… (2) X = {(Y-YM) / BB} + XM ………………………… (3 ) S3 (Determination of tip circle position) Next, in order to determine the tip circle position, a straight line that is parallel to the obtained regression line and intersects the tip circle position is obtained. In FIG. 8, the regression line is the visual field boundary (Y = 0 to 447, X = 0 to 511).
Positions [LX1, LY1] and [LX2, LY] that intersect with
2] is LX1 = XM- (YM / BB) ………………………… (4) LY1 = 0 ………………………… (5) If LX1 < If 0 or LX1> 511, LX1 = 0 LY1 = YM-BBT × XM LX2 = {(447-YM) / BB} + XM (6) LY2 = 447 …………………… (7) If LX2 <0 or LX2> 511, LX2 = 511 LY2 = YM + {BB × (511-XM) However, the ridge line is parallel to the Y axis and the machine size is adjusted so that it becomes the center of the visual field. Therefore, (4), (5) and (6), (7) are usually adopted.

Then, points [LX1-50, LY1] and (LX2-5) are added to the binary image used for the edge extraction.
0, LY2) (regression line in the horizontal axis is "-5
A pixel value distribution on a straight line that is translated by "0") is obtained.
This is to ensure that the straight line passes through both ends of the binary image in the Y-axis direction, that is, the tip circles, and the amount 50 of parallel translation in the X-axis direction is determined in advance from mechanical dimensions.

As shown in FIGS. 9A and 9B, FIG.
Point E at which the pixel value distribution in (A) changes from "1" to "0"
The Y-axis coordinate value of is the tip circle position. S4 (Setting of Reference Line) Next, as shown in FIG. 10, a linear function representing the pitch circle is calculated from the tip circle position by the following procedure. In FIG.
Since the distance T0FT from the tip circle position to the pitch circle is clear from the mechanical dimension in advance, it can be converted into coordinate values based on the magnification of the condenser lens 8. Further, the distance LCT from the visual field boundary parallel to the X axis to the position of the tip circle can be similarly converted into coordinate values. Therefore, the distance LCTP from the visual field boundary parallel to the X axis to the pitch circle can be calculated by LCTP = LCT + T0FT (8).

Subsequently, as shown in FIG. 11, in order to obtain the linear function of the reference line orthogonal to the regression line of the estimated ridge line, first, only the distance LCTP from the point [LX1, LY1] on the regression line on the regression line. A distant point, that is, an intersection [LCY, LCX] between the reference line to be obtained and the regression line is a distance L
It is obtained by a geometric operation of a right triangle having CTP as the length of the hypotenuse.

Since the slope BB of the regression line is the ratio between the perpendicular and the base of the right triangle, the intersection [LCY, LCX] is
When BB> 0, LCX = {LCTP / (BB ² +1)} ^1/2 + LX1 ………… (9) LCY = (LCX-LX1) × BB + LY1 ………… (10) BB <0 At some time, LCX = − {LCTP / (BB ² +1)} ^1/2 + LX1 (11) LCY = (LCX-LX1) × BB + LY1 (12)

Next, the coordinates [LSX1, LSY1] and [LSX2, LSY2] at which the reference line intersects the visual field boundary in the Y-axis direction.
Is LSX1 = BB × LCY + LCX (13) LSY1 = 0 ... (14) If LSX1 <0 or LSX1> 511, LSX1 = 0 .... (13) 'LSY1 = LCX / BB + LCY (14)' and LSX2 = -BB * (447-LCY) + LCX (15) LSY2 = 447 .... (16) If LSX2 <0 or LSX2> 511, then LSX2 = 0 ... (15) 'LSY2 = (-511 + LCX) / BB + LCY ... (16)' However, , The ridge is parallel to the Y-axis and the mechanical dimensions are set so as to be the center of the field of view, so (13) ′, (1
4) ′, (15) ′, and (16) ′ are adopted.

In this way, a straight line connecting the coordinates [LSX1, LSY1] and [LSX2, LSY2] is obtained as the reference line. S5 (determination of ridge line position and left and right tooth flank contour positions) Next, the ridge line position passing on the reference line is indicated by 2 in FIG. 9 (B).
Determined from the value image (single-sided image).

As shown in FIG. 12A, this is performed by sampling a one-sided image along the reference line and
A pixel value distribution as shown in is obtained. Then, the position where the pixel value changes from “0” to “1” in the obtained pixel value distribution is set as the ridge line position CPCL. Further, the pixel value distribution on the reference line is obtained for the smoothed multi-tone grayscale image processed in step S12 (FIGS. 13A and 13B).

The obtained pixel value distribution is shown in FIG.
As shown in (B), the minimum pixel value in the range [0, CPCL] is VL1, and the pixel value in the range [CPCL, 511] is VL2. Further, the offset value "30" is added to the larger one of VL1 and VL2 to generate the threshold value VL.

VL = max (VL1, VL2) +30 (17) The offset value “30” is an empirical value. Next, using the above VL as a threshold value, an image obtained by binarizing the multi-gradation grayscale image of FIG. 13A, that is, a binary image displaying the double-sided surface shown in FIG. 14A is generated. To do. 2 of this double-sided surface
The pixel value distribution on the reference line is calculated for the value image (see FIG. 14).
(B) The position where the pixel value changes from "1" to "0" is the left tooth flank contour position CPL, and the position where the pixel value changes from "0" to "1" is the right tooth flank contour position CPR.

S6 (Calculation of distribution amount) The distribution amount RATMAT is calculated by the following equation. RATEMAT = (CPR-CPCL) / (CPCL-CLP) ... (18) However, CPL <CPCL <CPR. S7 (Distribution Amount Judgment) The calculated distribution amount is passed or rejected according to the following procedure.

A = CPR-CPCL, b = CPCL-C
Assuming LP, the following relationship holds between the drawing tolerance t and a and b. b = a−t However, when a <b, t> 0.0, and when a> b, t <
0.0. When the distribution amount formula is expressed by a, b and t, RATEMAT = a / b = 1 / b × (b + t) = 1 + t / b (19) Here, the measured value t = 0.2 [mm], b = 1.83 [m
m] is substituted, t / b = 0.109≈0.11. Therefore, if 0.89 <RATMAT <1.11 holds, it is determined that the distribution amount is abnormal (NG).

The above determination is performed for all teeth of one piece, and defective products can be accurately sorted. As described above, in one embodiment of the present invention, one of the double-sided surfaces is lit up and the single-sided image is extracted accurately and clearly, and a regression line that estimates a ridge line from the single-sided image is obtained. Since the left and right tooth flank contour positions are determined from the regression line, the distribution amount can be calculated extremely accurately, and the accuracy of the success or failure of the distribution amount determination is improved.

The embodiment described above is an example, and the present invention can be variously modified without departing from the spirit of the claims. That is, even if the brightness of the double-sided surface is uniform, the image processing for extracting the single-sided image is possible. Further, as the reference line orthogonal to the regression line, the pitch circle is adopted in the embodiment, but the present invention is not limited to this, and the reference line is any if it intersects with the ridge line inside the tooth from the tip circle position. It may be a line set at the position. Therefore, the distance T0FT shown in FIG. 10 can be set to any value within the tooth height.

Further, the tip circle position determining means is based on the binary image used for extracting the tip circle position in the ridge line extraction, that is, along the straight line obtained by translating the regression line by "-50" in the horizontal axis direction. The above binary image is sampled and the tip circle position is obtained from the pixel value distribution based on the sampled image. However, the captured image is sampled along the regression line without using such a parallel-moved straight line, Obtaining the tip circle position from the pixel value distribution does not affect the accuracy of the pass / fail judgment.

[Brief description of drawings]

FIG. 1 is a flow chart showing an invention according to claim 1.

FIG. 2 is an explanatory diagram illustrating a chamfered surface measuring device for a gear and a chamfered surface illumination method according to an embodiment of the present invention.

FIG. 3 is a flowchart showing the operation of the embodiment of the present invention.

FIG. 4 is a screen view (A) and a pixel value distribution view (B) thereof showing a picked-up image of one tooth obtained by illuminating with different brightness.

FIG. 5 is a screen view showing a binarized image of a captured image.

6 is a screen view obtained by extracting a region having the largest area from the binary image of FIG.

FIG. 7 is an explanatory diagram showing a process of calculating a regression equation of an estimated edge line.

FIG. 8 is an explanatory diagram showing a regression line of an estimated edge line.

FIG. 9 is an explanatory diagram showing a process of determining the tip circle position.

FIG. 10 is an explanatory diagram showing a relationship of a reference line with respect to a tip circle position.

FIG. 11 is an explanatory diagram showing how to obtain a reference line orthogonal to a regression line.

FIG. 12 is an explanatory diagram showing a process of sampling a single-sided image with a reference line and determining a ridge line position from the obtained pixel value distribution.

FIG. 13 is an explanatory diagram showing a manner in which an image of a double-sided surface is binarized as a pre-stage of determination of left and right tooth flank positions.

14 is an explanatory diagram showing a process of determining left and right tooth flank positions from a binary image of a double-sided surface obtained by the process of FIG.

FIG. 15 is an explanatory view showing the operation of the synchronizer ring and the sleeve.

FIG. 16 is a perspective view showing a gear piece having a double-faced surface.

FIG. 17 is an explanatory diagram illustrating a distribution amount of a double-sided surface.

[Explanation of symbols]

1 is a gear piece to be inspected, 2 is a closed box, 3 is a jig, 4 is illumination means, 5 is imaging means, 8 is a condenser lens, 10 is an image processing device, 11 is a host computer, and 12 is a communication control device. .

Claims (3)

[Claims] 1. A double-sided chamfer is processed so that a ridgeline is formed in a direction normal to a pitch circle on an end surface of each tooth of a gear,
A chamfered surface measuring device for a gear, which measures a ridgeline position of the processed double-sided surface to measure a distribution amount of the double-sided surface, comprising: an image pickup means for picking up the double-sided surface; A single-sided chamfer extraction means for extracting a single-sided image representing a single-sided chamfer from the obtained captured image, and an estimation estimated in a direction substantially orthogonal to the single-axis based on a plurality of parallel uniaxial pixel value distributions of the single-sided image Regression formula setting means for obtaining the regression formula of the ridge line, addendum circle position determining means for obtaining the addendum circle position based on the regression line defined by the regression formula, and addendum determined by the addendum circle position determining means A reference line setting unit that sets a reference line orthogonal to the regression line inward by a predetermined distance from a circle position, and a pixel value distribution of the captured image in the direction of the reference line set by the reference line setting unit. Therefore, the distribution amount determination point calculation means for determining the pixel value change point representing the ridge line position and the two pixel value change points representing the left and right tooth flank contour positions from the pixel value distribution, and the distribution amount determination point calculation means. Chamfering of gears, comprising: computing means for computing the distribution amount from the respective pixel value change points; and determining means for deciding the quality of the gear by processing the chamfered surface based on the result of the computing means. Surface measuring device. 2. The one-side chamfered surface extracting means is the one-sided image as a binary image in which a maximum area area of the same pixel value of the captured image excluding the background portion obtained by the image capturing means and another area are divided into shades. The chamfered chamfering device for gears according to claim 1, wherein the chamfered surface measuring device is a maximum area extracting means for extracting an image. 3. A double-sided chamfer is processed so that a ridgeline is formed in a direction normal to a pitch circle on an end face of each tooth of a gear,
A chamfered surface illumination method of a gear that illuminates the double-sided surface when measuring the ridgeline position of the processed double-sided surface to measure the distribution amount of the double-sided surface, and capturing the double-sided surface. Image pickup means, and a region having the maximum area of the same pixel value of the imaged image obtained by the image pickup device and another region are shaded in order to extract a single-sided image showing a single chamfered surface from the imaged image obtained by the image pickup device. A single-sided chamfer extraction means for extracting the single-sided image as a binary image divided into two, and regression of an estimated ridge line estimated in a direction substantially orthogonal to the single-axis based on a plurality of parallel uniaxial pixel value distributions of the single-sided image. Regression formula setting means for obtaining a formula, addendum circle position determining means for obtaining addendum circle position based on a regression line defined by the regression formula, and addendum circle position determined by the addendum circle position determining means A reference line setting means for setting a reference line orthogonal to the regression line inward by a predetermined distance, and a pixel value distribution of the captured image in the direction of the reference line set by the reference line setting means, Distribution amount determination point calculation means for determining a pixel value change point representing a ridge line position and two pixel value change points representing both contour positions from the pixel value distribution, and each pixel value change determined by the distribution amount determination point calculation means Using a chamfered surface measuring device for a gear, which comprises a calculating means for calculating the distribution amount from a point and a judging means for judging whether the gear is good or bad on the basis of the result of the calculating means, A method for illuminating a chamfered surface of a gear, which comprises illuminating so as to cause a difference in brightness.