JPH0453254B2 - - Google Patents

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a technology for determining the quality of repeating patterns of the same shape using autocorrelation images, and is concerned with the technology for determining the quality of repeating patterns of the same shape by using autocorrelation images. , relates to technology used as an inspection means for detecting scratches, abnormal irregularities, etc.

[Prior Art] This invention is an invention that applies the invention related to the earlier application of the same applicant (the "correlation device" of Japanese Patent Application No. 58-191691 (Japanese Unexamined Patent Publication No. 60-84666)).

[Problem that the invention seeks to solve]

There is a strong desire to inspect the core material of optical fibers for bends, scratches, irregularities, etc. during the assembly line production process, but the appropriate technology to meet this requirement remains lacking. In particular, the current situation is that it is not possible to recognize patterns in assembly lines and compare them with reference patterns of good products. This invention has been made to provide a means for solving these problems.

[Means for solving problems]

In order to solve this problem, the present invention utilizes two characteristic facts.

First, in order to detect bends, scratches, irregularities, etc. in the core material of an optical fiber, we focused on the fact that the grooves in the core material have a spirally rotated structure with a certain pitch. A parallel beam of light is irradiated onto the object, and the shadow pattern created by the transmitted light is captured and imaged with a television camera. This shadow picture pattern can be obtained with regular good quality core materials, but
This is a fact that clearly reflects the presence of dust inside the grooves, irregularities in shape caused by scratches, or irregularities in the core material.

Second, in order to compare the pattern obtained by irradiation with a parallel light beam with a reference pattern of a good product during the assembly process, a correlator operating in real time generates an autocorrelation pattern. The autocorrelation pattern has the characteristic that, for a moving object, it appears as a motionless pattern with fixed periodicity and shape characteristics (also known as the disappearance of the phase component in autocorrelation calculations). It is also possible to fix and observe the movements of objects on assembly lines. Therefore, it becomes easy to compare the fixed autocorrelation pattern with the autocorrelation pattern of a non-defective product and perform an inspection to determine whether the product is good or bad.

〔Example〕

FIG. 1 shows a configuration diagram of an example in which the method of determining the quality of repeating patterns of the same shape according to the present invention is implemented as an inspection apparatus for the core material of an optical fiber. As shown in the figure, the core material 1 of the object to be measured has a spiral shape in the length direction, and assembly work is being performed in which the core material 1 moves in the direction of the longitudinal arrow 2. The optical pattern detector 3 is a device for obtaining a repeating pattern of the same shape by irradiating a parallel light beam onto an article to be measured (or an article to be referred to as a reference) to determine the quality. It consists of an optical system that uses a laser as a light source and a television camera that takes a shadow picture of the core material. A core material, which is an object to be measured (or a reference object), moves in the direction of the arrow in this detector, and the obtained television image is sent to a correlator (preferably an acousto-optic correlator) 4 that operates in real time. The correlator can directly input the television signal obtained by scanning a pattern in which the same shape is repeated in the repeating direction of the parallel light beam in the pattern repeating direction, and can also directly input the television signal obtained by scanning the pattern in the repeating direction of the pattern, which is created by parallel light beams.Moreover, it is possible to directly input the television signal obtained by scanning the pattern in which the same shape is repeated, which is created by parallel light beams. An output signal can be output. Regarding the autocorrelation image obtained from the reference article, feature points of the reference article may be extracted and stored in advance according to the brightness of the image at each position. This memory is stored in the feature point memory 5. Regarding the autocorrelation image obtained from the article to be judged as good or bad, feature points are extracted according to the brightness of the image at each position, and are compared with the stored contents of the feature point memory 5 in the feature point comparator 6,
The quality determination result as to whether or not a change is observed is displayed on the display 7.

FIG. 2 shows a specific example of the structure of the optical pattern detector 3, in which a is a top view and b is a front view. He
- The light beam from the Ne laser light source 31 is expanded into a parallel beam by two lenses 32 and 33, and a filter 34 necessary for smoothing the brightness of the beam is placed in the groove of the core material 1 of the object to be measured. The core material is irradiated so that the direction of the parallel light beam almost coincides with the direction of the parallel light beam. By projecting light from this direction, the core material creates a repeating pattern of the same shape. A repeating pattern of the same shape that is partially shielded by the core material is focused by a lens 35, and a pattern projection image of the light beam is obtained by an image pickup tube 36 of a television camera. This pattern of light flux is a silhouette of the grooves in the core material, and is obtained as an electrical signal by scanning the projected image in the direction in which the pattern is repeated.

[Effect]

FIG. 3 shows the results of detecting a projected pattern image from the groove of the core material of the optical fiber using the optical pattern detector shown in FIG. The difference between a and c is whether the angle θ at which the parallel light beam is irradiated was adjusted to the pitch of the core material core (see figure b), or the pattern was substantially enlarged by increasing θ a little. However, a silhouette with the same pattern repeated in the scanning direction of the television is obtained.

FIG. 4 is a pattern projection image detected by the optical pattern detector for a core material having a pitch deviation. It is observed that the pitch of the groove at the right end of the two objects to be measured, a and b, is large.

FIG. 5 shows projected patterns b and c of the grooves of the optical fiber when dust is attached, side by side with a normal one without dust. Irregular shapes of grooves due to debris are observed.

FIG. 6 is a diagram illustrating a procedure for adding an electrical signal obtained by an optical pattern detector to a correlator, transforming it into an autocorrelation image, and extracting feature points by the position of brightness. . FIG. 3A is a pattern projection diagram of a normal core material corresponding to FIG. 3, in which bright parts that are not directly related to the shape of the grooves are shaded. The autocorrelation image of this normal core material is as shown in figure b. The scanning direction is the direction in which the pattern is repeated, that is, the direction of the arrow in Figure A, and the correlation definition section is set to fill the entire width of the screen. As is clear from Figure b, the correlation output screen naturally appears symmetrically, and this image remains unchanged despite the horizontal movement of the input image, which is a feature of the autocorrelation image. However, if image deformation such as pitch shift occurs in the input image, the autocorrelation image will also be deformed. Experimentally confirm the part where this deformation is most noticeable, and check whether a signal always appears at that part (position) in the projection view from a normal object.
The criterion for determining the characteristics is whether or not it appears. For example, in figure b, the ○ marks are feature points with signals, and the ● marks are feature points with no signals. The position of this feature point and information on the presence or absence of a signal can be easily stored in memory. An example of a case where the object to be measured has a deviation in pitch (corresponding to FIG. 4) is similarly shown in Fig. c for a pattern projection diagram, and Fig. d for an autocorrelation image. The same applies to ○ marks and ● marks. In figure d, the ● mark position, which originally has no signal, has brightness, which is different from the reference drawing pattern, and the ○ mark position does not have brightness, which is also different from the reference figure pattern, indicating a shape abnormality. It is determined that

〔Effect of the invention〕

As explained above, in this invention, a repeating pattern of the same shape is obtained by irradiating a parallel light beam, and then an autocorrelation image of the pattern is obtained by a correlator operating in real time, and each position of the autocorrelation image is Since the feature points of the reference article and the article to be measured are extracted and compared according to the luminance of Now we can measure things in the flow. Furthermore, we were able to skillfully utilize two features of autocorrelation images: invariance of position and clear appearance of feature points.

[Brief explanation of the drawing]

FIG. 1 shows a basic configuration diagram of an embodiment of a core material inspection apparatus using the method of the present invention. Figure 2 is the first
FIG. 2 is a diagram showing the configuration of an optical pattern detector used in the embodiment shown in the figure, in which figure a shows a top view and figure b shows a front view. FIG. 3 is a pattern diagram of grooves in a normal core material detected by an optical pattern detector, a and c are projection diagrams of the pattern, and b and d are diagrams for explaining the projection situation. FIGS. 4a and 4b both show projected patterns of grooves in the core material with abnormalities. FIG. 5a shows a projected pattern of grooves in a normal core material, and FIG. 5b and c show projected patterns of grooves with dust attached, respectively. FIG. 6 shows a procedure for applying the electrical signal obtained by the optical pattern detector to a correlator, transforming it into an autocorrelation image, and extracting feature points according to their brightness positions. In the figure, 1 is the core material of the object to be measured, 2 is the moving direction, 3 is the optical pattern detector, 4 is the correlator, and 5
6 is a feature point memory, 6 is a feature point comparator, and 7 is a display device.

Claims (1)

[Claims] 1. Obtaining a projected image of a repeated pattern of the same shape by irradiating a reference article with a parallel light beam onto a reference article having a repeated pattern of the same shape; scanning the projected image in the direction in which the pattern is repeated and converting it into an electrical signal. and inputting it into a correlator to obtain a reference autocorrelation image; storing the positions of the feature points of the reference article according to the brightness at each position of the reference autocorrelation image; and determining whether it is good or bad. A step of irradiating the object with a parallel light beam to obtain a projected image of a repeating pattern of the same shape; scanning this projected image in the direction in which the pattern is repeated and converting it into an electrical signal, inputting it to a correlator and transmitting it to the judgment self. a step of obtaining a correlation image, detecting the position of a feature point of the judgment article according to the brightness at each position of the autocorrelation image for judgment, and comparing the position of the feature point with the stored position of the feature point of the reference article; A method for determining the acceptability of repeating patterns of the same shape, which comprises the steps of comparing.