Gear inspection method

Abstract

To provide a gear inspection method capable of inspecting the shape of a tooth surface with high accuracy.SOLUTION: A gear inspection method of the present invention is an inspection method for detecting abnormalities in the shape of the tooth flank of a gear W, and includes the steps of imaging the tooth flank to obtain an image, obtaining from the image a brightness value for each pixel on a pre-specified line on the tooth flank, and detecting abnormalities in the shape of the tooth flank from the transition of the brightness values of the pixels on the line.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a method for inspecting gears.

Background Art

[0002] In order to prevent the outflow of defective products, it is necessary to inspect the shape of a formed gear. As a method for inspecting a gear, there is a method of inspecting using a gear measuring machine. The tooth surface of the gear is traced by the contact probe of the gear measuring machine, and the accuracy of the gear (tooth profile, tooth flank, pitch, runout of tooth groove, etc.) is inspected. However, inspection by a gear measuring machine takes time. Therefore, as a method for inspecting a gear, a method has been proposed in which a gear is imaged by a camera and it is detected whether there is an abnormality in the shape of the gear from the captured image. For example, in the gear inspection method described in Patent Document 1, defects of the gear are detected based on difference data between an image of the gear captured and an image of a non-defective product stored in advance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Regarding the gear inspection method disclosed in Patent Document 1, the inventor has found the following problems. In the gear inspection method disclosed in Patent Document 1, the shape of the tooth surface is extracted from the captured image, and the size of the region detected as a defect is evaluated. Therefore, with this inspection method, it is possible to detect clear uneven abnormalities on the tooth surface such as dents and scale remnants, but it is difficult to perform a high-precision inspection such as detecting small uneven changes when looking at the entire shape.

[0005] The present invention has been made to solve such problems, and provides a gear inspection method capable of inspecting the shape of a tooth surface with high precision.

Means for Solving the Problem

[0006] The gear inspection method according to the present invention is an inspection method for detecting an abnormality in the shape of the tooth surface of a gear, comprising: a step of imaging the tooth surface to obtain an image; a step of obtaining the luminance value for each pixel on a line specified in advance on the tooth surface from the captured image; and a step of detecting an abnormality in the shape of the tooth surface from the transition of the luminance values of the pixels on the line.

Advantages of the Invention

[0007] According to the present invention, it is possible to provide a gear inspection method capable of inspecting the shape of a tooth surface with high precision.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Modes for Carrying Out the Invention

[0009] Embodiment 1 Hereinafter, with reference to the drawings, the gear inspection method according to Embodiment 1 of the present invention will be described. In this embodiment, as a method for inspecting the shape of the tooth surface of a gear, first, the tooth surface is imaged by a camera, and the luminance value for each pixel is obtained from the image. Then, an abnormality in the shape of the tooth surface is detected from the transition of the luminance values.

[0010] <Configuration of Gear Inspection Apparatus> First, with reference to FIG. 1, the configuration of an inspection apparatus for realizing the gear inspection method according to Embodiment 1 of the present invention will be described. The inspection device 10 in the first embodiment includes a gear rotation mechanism 1, a camera 2, and a computer 3.

[0011] The gear rotation mechanism 1 rotates the gear W, which is the object to be inspected, for imaging by the camera 2. The gear rotation mechanism 1 rotates the installed gear W according to the rotation speed and rotation time input by a control device (not shown). The gear rotation mechanism 1 includes, for example, a motor, various gears, and a motor control mechanism.

[0012] The camera 2 is an imaging means used to image the tooth surface of the gear W. In this embodiment, as the camera 2, for example, a line camera capable of imaging the entire circumference of the side surface of the gear W as a single image is preferable. The line camera images the object linearly and generates a two-dimensional image by synthesizing the captured one-dimensional images. There are two types of line cameras: monochrome and color. In this embodiment, a monochrome line camera is preferable from the viewpoint of obtaining luminance values from the captured tooth surface image and further performing preprocessing of the data.

[0013] The computer 3 is a control means including an image processing unit 4 and an image determination unit 5. The computer 3 is composed of, for example, a CPU (Central Processing Unit) and various memories. In this embodiment, the image processing unit 4 and the image determination unit 5 are provided in the same computer 3, but they may be provided in separate computers 3.

[0014] The image processing unit 4 receives the image acquired by the camera 2 and executes processing of the image. Then, the data obtained from the image processing is output to the image determination unit 5. The data obtained by processing the image is the luminance value for each pixel. The processing of the image will be described in detail later.

[0015] The image determination unit 5 detects abnormalities in the shape of the tooth surface based on the data output input by the image processing unit 4. The image determination unit 5 has previously read in data of good products and performed machine learning. That is, the image determination unit 5 generates a learned prediction model by executing machine learning. Then, the image determination unit 5 uses the generated prediction model to determine whether the tooth surface reproduced by the inspection data is abnormal. In addition, the image determination unit 5 outputs the determination result of the tooth surface to a notification unit such as a display (not shown).

[0016] <Gear inspection method> Subsequently, with reference to FIG. 2, the gear inspection method according to Embodiment 1 will be described.

[0017] First, the gear rotation mechanism 1 rotates the gear W installed on the gear rotation mechanism 1 at a predetermined rotation speed (step S101). In a typical example, the gear rotation mechanism 1 rotates the gear W at a constant rotation speed. The gear rotation mechanism 1 continues to rotate until imaging of the gear W is completed. Next, the camera 2 images the tooth surface of the gear W (step S102). The camera 2 transmits the captured image to the image processing unit 4.

[0018] Next, the image processing unit 4 receives from the camera 2 the image of the tooth surface captured by the camera 2, and extracts an image of the tooth surface for one tooth (step S103). The extraction of the image of the tooth surface for one tooth is performed by cutting out the image of the tooth surface for each tooth and obtaining as many images as the number of teeth imaged.

[0019] Next, the image processing unit 4 acquires data of luminance values for each pixel from the image of the tooth surface for one tooth as the shape of the tooth surface (step S104). The image processing unit 4 designates a location to be inspected on the captured image with a straight line or a curve, i.e., a line. Then, the image processing unit 4 acquires data of luminance values for each pixel on the line designated on the tooth surface. As the direction of line designation, for example, it is preferable to designate the cutting direction of the tooth surface or a direction orthogonal to the cutting direction. By designating these directions, abnormalities in the tooth streaks (the shape in the cutting direction of the tooth surface) and tooth profiles (the shape in the direction orthogonal to the cutting direction of the tooth surface) can be detected. Also, the image processing unit 4 may execute image processing before acquiring the luminance value for each pixel. As the image processing, for example, it is preferable to normalize the brightness of the image.

[0020] Next, the image processing unit 4 executes preprocessing of the data (step S105). As a method of preprocessing, for example, there is a method of taking a moving average of luminance values for each pixel, but it is not limited to this. The image processing unit 4 outputs the preprocessed data to the image determination unit 5. By performing preprocessing of the data, the accuracy of detecting abnormalities on the tooth surface by the image determination unit 5 can be improved.

[0021] Finally, the image determination unit 5 detects an abnormality in the shape of the tooth surface based on the data output input by the image processing unit 4, and determines whether the gear W is a good product or a defective product (step S106). The image determination unit 5 uses a prediction model generated in advance by machine learning to determine whether there is an abnormality in the data output input by the image processing unit 4. Then, the image determination unit 5 outputs the determination result of the tooth surface to a notification unit such as a display (not shown).

[0022] <Example> Hereinafter, an example of the inspection method for a gear according to the first embodiment will be described. In the example, a hypoid gear was used as the gear W to be inspected. Also, in step S104, data of luminance values for each pixel were acquired along a direction orthogonal to the cutting direction of the tooth surface as the inspection target direction. Therefore, abnormalities in the tooth streaks (the shape in the cutting direction of the tooth surface) of the gear were detected.

[0023] FIG. 3 shows the data of the luminance value for each pixel determined in step S106. FIG. 3(a) shows the data of the tooth surface determined as a good product, and FIG. 3(b) shows the determination data of the tooth surface determined as a defective product. As shown in FIG. 3, for example, the convex scratches on the tooth surface appear at the locations where the luminance value in the data of the defective product is larger than the luminance value in the data of the good product. Also, the dents on the tooth surface generated due to deep processing on the tooth surface appear at the locations where the luminance value in the data of the defective product is smaller than the luminance value in the data of the good product. Thus, the convex scratches and the dents due to excessive processing on the tooth surface can be discriminated from the transition of the luminance value for each pixel.

[0024] As described above, according to the method for inspecting a gear according to the present embodiment, the tooth surface is imaged by a camera, and the abnormality of the shape of the tooth surface can be detected in a short time and with high accuracy from the transition of the luminance value for each pixel obtained from the image of the tooth surface. Furthermore, it has become possible to detect abnormalities in the tooth streaks (the shape in the cutting direction of the tooth surface) and the tooth profile (the shape in the direction orthogonal to the cutting direction of the tooth surface), which were difficult to detect by inspection by image determination heretofore.

[0025] Note that the present invention is not limited to the above-described embodiment, and can be appropriately modified without departing from the gist thereof.

Description of Reference Numerals

[0026] 1 Gear rotation mechanism 2 Camera 3 Computer 4 Image processing unit 5 Image determination unit 10 Inspection device W Gear

Claims

【Claim 1】 An inspection method for detecting an abnormality in the shape of a tooth surface of a gear, comprising: imaging the tooth surface to obtain an image; obtaining a luminance value for each pixel on a line specified in advance on the tooth surface from the imaged image; and detecting an abnormality in the shape of the tooth surface from the transition of the luminance values of the pixels on the line. A gear inspection method.

Images