JPH04304576A - Picture processor - Google Patents

Abstract

PURPOSE:To improve the accuracy of identification as against a specified face by always and correctly picking up the presence or absence, position and a shape, etc., of a projecting and recessing part in the specified face of an object. CONSTITUTION:This picture processor is constituted of plural light sources 2 or 4 which diagonally illuminate the specified face from respectively different directions, CPU28 which lighting-controls the plural light sources in accordance with plural lighting patterns, a picture memory 23 which individually stores plural different picture data obtained under illumination formed by the lighting control and a intra-picture operating part 32 which executes an intra-picture operation based on respective picture data stored in the memory 23 so as to be constituted to identify the specified face based on the operation result of the operating part 32 and to output the identified result.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, and particularly to an image processing apparatus that illuminates and images an object to be inspected passing through an inspection surface in a production line in a factory, and uses the result of this imaging to take an image of the object. The present invention relates to an image processing device that performs identification such as quality determination processing of inspection objects.

0002

[Prior Art] In an inspection line in a factory where an image processing device is used, when the object to be inspected is inspected based on engraved characters or concavities on the surface such as scratches, the following conventional method is used. The test was conducted in a similar manner.

[0003] As a first method, uniform reflected illumination is applied to the object to be inspected from the same direction as the imaging direction of a television camera, which is an imaging device, using a round fluorescent lamp or the like. The image signal obtained by imaging under this illumination condition is binarized using a predetermined threshold value, and the presence/absence and position of a shadow formed by a concave portion on the surface of the object to be inspected is determined.
There is a method of extracting the shape, etc., and determining whether the inspected object is good or bad based on the extraction result.

[0004] As a second method, a straight tube fluorescent lamp or the like is used to provide oblique illumination to the object to be inspected. The image signal obtained by imaging under this illumination condition is binarized using a predetermined threshold value, and the presence/absence, position, shape, etc. of shadows caused by concave parts of the surface are extracted. There is a method for determining the quality of the object to be inspected based on the extraction result. In particular, with this method, the angle of incidence of the illumination light on the object to be inspected can be made shallow, so even if the concave portion is shallow,
This has the advantage that the contrast of shadows can be improved.

[0005]

However, the first method described above has a problem in that the determination results vary depending on the size and depth of the concave portion on the surface of the object to be inspected. That is, when the recessed portion is shallow, the contrast of the shadow becomes poor, and when the width of the recessed portion is so large that it cannot be ignored compared to the depth, the bottom of the recessed portion may be imaged brightly. Therefore, when attempting to binarize the obtained image signal, it is not possible to correctly extract the presence, position, and shape of the concave portion.
There is a problem in that it is not possible to accurately determine whether the inspected object is good or bad.

Furthermore, in the second method described above, since illumination is applied from one direction, there are areas where shadows are formed and areas where shadows are not formed depending on the direction of the recessed portion. Therefore, when the image signal obtained by imaging is binarized using a predetermined threshold, the position and shape of the concave portion cannot be extracted correctly, making it difficult to accurately determine pass/fail. There were also problems.

Therefore, it is an object of the present invention to always correctly extract the presence, position, shape, etc. of concave or convex portions on a predetermined surface of an object, thereby improving the accuracy of identification of the object to be inspected. An object of the present invention is to provide an image processing device.

[0008]

An image processing apparatus according to the present invention is an image processing apparatus that processes an image signal obtained by imaging a predetermined surface of a subject. In detail, the processing device includes a plurality of light sources that obliquely illuminate the predetermined surface from different directions, and individually illuminates the plurality of light sources according to a plurality of predetermined lighting patterns and sequentially updates the lighting patterns. For each lighting control means that controls lighting and the lighting pattern updated by the lighting control means,
a storage means for converting the image signal obtained by the imaging into image data and storing the image data individually; and a calculation means for performing an inter-image calculation based on each of the image data stored in the storage means;
The camera is configured to include identification means for identifying the subject based on the calculation result of the calculation means, and output means for outputting the result of identification by the identification means.

[0009]

[Operation] The image processing apparatus according to the present invention is configured as described above, and oblique illumination is applied from different directions to a predetermined surface of a subject for each lighting pattern of the light source that is sequentially updated by the transfer control means. is applied to obtain a plurality of different image data. Therefore, since the inter-image calculation by the calculation means is performed based on the plurality of image data, the accuracy of the calculation result becomes high, and the accuracy of the identification of the subject performed by the identification means based on the calculation result. It also improves.

0010

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

[0011] This embodiment will be described assuming that the image processing apparatus is applied to print inspection of engraved characters on an object.

FIG. 1 is a diagram showing a schematic configuration of an image processing apparatus according to an embodiment of the present invention.

In the figure, the image processing apparatus includes a camera 1 for taking an image of an object to be inspected with predetermined characters engraved on its surface, and a video signal outputted from the camera 1 inputted thereto, and image data processing is performed accordingly. Data processing unit 5 and first
to third light sources 2 to 4 and a photoelectric switch 33.

The image data processing section 5 includes the image input section 2
2. An image memory 23 which is a digital memory, an image output section 24, a video monitor 25, a timing control section 26, an address/data bus 27, a CPU (central processing unit) 28,
ROM29, RAM30, I/O (input/output) control unit 31
and an inter-image calculation section 32.

The image input unit 22 inputs an image captured by the camera 1 connected to the previous stage as an analog video signal, and accordingly A/D (analog/digital) converts the video signal to create a grayscale image. It is output as data to the image memory 23 connected to the next stage. In addition, the image input section 22
has a function that detects the amount of deviation from a predetermined reference position for the image obtained by imaging using a positional deviation detection window that is set to surround the edge of the object to be inspected on the image. Have both.

The image memory 23 has memory areas ma and mb that can store the given grayscale image data in units of one pixel and individually store image data for three screens.
and mc. Furthermore, data is read and written to and from the image memory 23 at arbitrary timing.

The inter-image calculation unit 32 selects two images from the image memory 23 based on the reading order determined based on the positional deviation amount calculated by the image input unit 22.
It has a function of reading out individual image data while correcting positional deviations, performing inter-image calculations, and storing the calculation results in the image memory 23 again.

The image output unit 24 inputs image data provided from the image memory 23 connected to the previous stage, performs D/A (digital/analog) conversion accordingly, and converts it into a video signal for one screen to be sent to the next stage. to a video monitor 25 connected to.

The video monitor 25 has a monitoring function,
Screen display is performed using the video signal supplied from the image output unit 24 connected to the previous stage.

An image memory 23, a CPU 28, a ROM 29, a RAM 30, and an I/O control section 31 are connected to the address/data bus 27, whereby a simple microcomputer is constructed by these sections. In this microcomputer, the CPU 28 executes a program stored in the ROM 29 in advance to perform various image processing such as feature calculation processing.

[0021] The timing control section 26
28, a timing signal for controlling data input/output in the image input section 22, image memory 23, and image output section 24 is generated and applied to each section.

FIGS. 2(a) and 2(b) are external views showing the installed state when the image processing apparatus shown in FIG. 1 is applied to the printing inspection of engraved characters.

In FIG. 2(a), the above-mentioned image processing device is provided to judge whether the shape of the character stamped on the surface of the object to be inspected 8 conveyed on the conveyor 6 is good or bad.

A predetermined inspection surface 7 is provided on the conveyor 6. When the inspection object 8 is transported and positioned on the inspection surface 7, it is detected by the photoelectric switch 33,
The detection signal is given to the image data processing section 5. Further, a camera 1 is fixedly provided so as to take an image of the object to be inspected 8 located on the inspection surface 7 from above. Further, three light sources 2 to 4 are fixedly provided diagonally above the inspection surface 7. As shown in FIG. 2(b), the light sources 2 to 4 have substantially the same incident angles 12, 13 and 14 of the light beams 9, 10 and 11 on the inspection surface 7, and the angles of incidence of the light beams 9 to 11 are approximately the same. Corners 15, 16 and 17 are fixedly provided on the inspection surface 7 to form an angle of 60°. These light sources 2 to 4 are
It has a lighting control function, such as a light source device with a shutter function.

FIGS. 3(a) to 3(d) are similar to FIG. 2(a) shown above.
FIG. 4 is a schematic diagram illustrating an example of marking of characters on the object to be inspected to be judged as pass/fail during the print inspection shown in FIGS.

FIG. 3(a) is a plan view showing an example of character engraving when the image processing apparatus determines that the product is non-defective 18. In the non-defective product 18, the characters ABC are clearly engraved on the stamp portion 19. A sectional view taken along the line AB in FIG. 3(a) above is shown in FIG. 3(b). As shown in FIG. 3(b), it can be seen that the stamped portion 19 of the non-defective product 18 is formed with the letters ABC as a concave portion on its surface. On the other hand, FIGS. 3(c) and 3(d) are plan views of the surfaces of the defective products 20 and 21. The defective product 20 shown in FIG. 3(c) is
The engraved letters ABC are in a defective engraved state with a missing character at the A part. Moreover, the defective product 21 shown in FIG. 3(d) has incorrect characters engraved on it.

The image processing device installed as shown in FIG. Regarding the part 19, if a defective product 20 or 21 with missing or incorrect characters is detected, it is output that the inspected object 8 is a defective product. It works like that.

FIG. 4 is a schematic process flow diagram showing the operation of inspecting the printing of engraved characters by the image processing apparatus provided as shown in FIGS. 2(a) and 2(b).

FIGS. 5A to 5E are schematic diagrams showing images obtained by executing the processing flow shown in FIG. 4 above.

Next, the operation of inspecting the characters engraved on the surface of the object to be inspected 8 of the image processing apparatus installed as shown in FIG. 2(a) will be carried out according to the processing flow shown in FIG. , will be explained with reference to FIGS. 1 to 3 and 5.

The processing flow shown in FIG. 4 is stored in advance as a program in a predetermined storage area of the memory RAM 30 and executed under the control of the CPU 28. In addition, each processing step of the processing flow shown in FIG.
The time required for one screen of image signals is assumed, and all processing within the same step is executed in the time for one screen.

As shown in FIG. 2(a), the conveyor 6
When the photoelectric switch 33 detects that the object 8 to be inspected, which has been transported by the . The CPU 28 starts executing the processing flow shown in FIG. 4 in response to the detection signal input.

At step ST1 in FIG. 4 (abbreviated as ST1 in the figure), the CPU 28 determines whether or not there is a detection input of the object to be inspected 8 from the photoelectric switch 33. At this time, the determination process in step ST1 is repeatedly executed until there is a detection input from the photoelectric switch 33.
When there is a detection input of 3, the CPU 28 executes the next step S.
Branches to processing after T2.

In the processing from step ST2 to step ST4, the CPU 28 first turns on the first light source 2, which is the first light source, and the video signal obtained by capturing the image from the camera 1 is inputted via the image input section 22. It is given to the image memory 23 as image data. The image memory 23 stores the image data in its memory area ma. Note that the image data obtained by turning on the first light source 2 will be referred to as image a. The image a represents an engraved character formed by unevenness as shown in FIG. 5(a). next,
The image input unit 22 calculates the positional shift amount Za of the image a.

The CPU 28 then executes the processing shown in steps ST5 to ST7. First, the first light source 2 is turned off, the second light source 3 is turned on, and the image data obtained in the same manner is stored in the memory area mb of the image memory 23 as image b in FIG. 5(b). . At this time, the positional deviation amount Zb of the image b is calculated in parallel.

Thereafter, the CPU 28 executes the processes shown in step ST8 and step ST10. first,
While the second light source 3 is turned off, the third light source 4 is turned off.
is turned on, and the obtained image data is stored in the memory area mc of the image memory 23 as image c in FIG. 5(c), and the positional shift amount Zc of the image c is calculated.

As described above, images a, b, and c are individually stored in memory areas ma, mb, and mc in the image memory 23, respectively. This image a to image c
5(a) to 5(c) each represent grayscale image data of stamped characters formed by unevenness as shown in FIGS. 5(a) to 5(c). As shown in FIG. 5(a), image a represents image data obtained under illumination conditions in which only light source 2 is turned on, and image b in FIG. 5(b) represents image data obtained under illumination conditions in which only light source 3 is turned on. Similarly, image c in FIG. 5C represents image data obtained under illumination conditions in which only the light source 4 is turned on.

Next, the CPU 28 executes inter-image calculation by the inter-image calculation unit 32 via the timing control unit 26 in steps ST11 to ST13. In other words, the inter-image calculation unit 32 uses the image memory 23
The images a and b stored in the memory areas ma and mb of
The image a and Zb are read out from the memory 23 while correcting the positional deviation, inter-image calculation is performed on both images, and the calculation result image ab is stored in the memory area ma.

The inter-image arithmetic operation executed by the inter-image arithmetic unit 32 is performed by mi
Assume that we are performing n operations. In response to the calculation result image ab being stored in the memory area ma, the CPU 28 controls the third light source 4 to turn off via the I/O control unit 31.

The calculation result image ab is stored in the memory area ma of the image memory 23 as image data having irregularities as shown in FIG. 5(d). It is clear that the image ab is a calculation result image obtained by performing a min calculation between the images in FIGS. 5(a) and 5(b).

After the calculation result image ab is calculated as described above, the CPU 28 subsequently executes the processing of steps ST14 to ST16 to calculate the image abc. Specifically, the inter-image calculation unit 32 performs the inter-image calculation using the image ab and the image c stored in the memory area mc of the image memory 23 in the same manner as described above. This calculation result image abc is stored in the memory area mb of the image memory 23. It can be seen that the calculation result image abc represents the min calculation result of the image ab of FIG. 5(d) and the image c of FIG. 5(c), as shown in FIG. 5(e).

Calculation result image a obtained as described above
Since the shadow portion formed by the individual illumination of light sources 2 to 4 is dark, bc is
It can be seen that the concave portion of the stamp is correctly extracted by executing the n operation.

[0043] The CPU 28 performs steps ST17 and S
By further binarizing the image abc in which the concave portions of the stamp were correctly extracted in the process of T18 using a predetermined threshold value, the concave portions stamped on the surface of the object to be inspected 8 can be extracted. can be extracted accurately. Then, the CPU 28 compares and checks the result of the binarization process with the binarized data serving as a predetermined reference, and determines whether the stamped characters formed on the surface of the object to be inspected 8 are good or bad. Then, the determination result is displayed on the video monitor 25 via the address/data bus 27, or
The data may be output to a display device or a printing device (not shown) via the O control unit 31. After that, the process returns to step ST1 again.

[0044] The quality determination process in step ST17 is performed by registering in advance the area value or image pattern of the non-defective product 18 in a region of interest set in a portion where a stamp is always present as a reference value, and A common method is to compare the measurement data of the inspection object 8 with the reference value to determine whether it is good or bad.

In the embodiment described above, three light sources 2 to 4 are used for illuminating the object 8 to be inspected, but the number of illuminating light sources is not limited to three; Increase the pattern by adding 2 or 4 or more,
The calculation accuracy may be improved. moreover,
The inter-image calculation in the inter-image calculation unit 32 may be executed by software processing by the CPU 28. Further, the inter-image calculation is not limited to the min calculation, and another calculation method that obtains a similar effect may be used.

[0046]

According to the image processing apparatus according to the present invention, it is possible to obtain a plurality of different image data by applying a plurality of oblique illuminations to a predetermined surface of a subject from different directions, so that calculations between images can be performed mutually. The accuracy of the calculation results improved, for example, even if the concave portion formed on the predetermined surface is shallow,
This has the effect of improving the contrast of shadows and correctly extracting the entire concave portion.

Furthermore, since the plurality of light sources for providing oblique illumination to the predetermined surface are individually arranged,
For example, there is an effect that the conveyance route such as a conveyor of an inspection line through which the subject is conveyed is not obstructed.

Furthermore, since a good calculation result image can be obtained within the time required for the same number of image frames as the number of oblique illuminations, it is possible to perform high-accuracy and high-speed identification processing on the predetermined surface. There is.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a schematic configuration of an image processing device according to an embodiment of the present invention.

2A and 2B are external views showing an installed state when the image processing apparatus shown in FIG. 1 is applied to a print inspection of engraved characters.

FIGS. 3(a) to (d) represent FIGS. 2(a) and (b).
) is a schematic diagram illustrating an example of marking of characters on an object to be inspected to be judged as pass/fail during the print inspection shown in FIG.

FIG. 4 is a schematic processing flow diagram showing the operation of inspecting the printing of stamped characters by the image processing apparatus provided as shown in FIGS. 2(a) and 2(b).

5A to 5E are schematic diagrams showing images obtained by executing the processing flow shown in FIG. 4;

[Explanation of symbols]

1 Cameras 2, 3, and 4 Light source 5 Image data processing section 22 Image input section 23 Image memory 25 Video monitor 28 CPU 32 Inter-image calculation section ma, mb, and mc Memory area Note that in each figure, the same reference numerals indicate the same or equivalent parts. shows.

Claims (1)

[Claims] 1. An image processing device that processes an image signal obtained by imaging a predetermined surface of a subject, comprising: a plurality of light sources that obliquely illuminate the predetermined surface from different directions; lighting control means for controlling the lighting of the plurality of light sources individually according to the plurality of lighting patterns and sequentially updating the lighting patterns; storage means for converting the image signals into image data and storing them individually; a calculation means for performing inter-image calculations based on the respective image data stored in the storage means; and based on the calculation results of the calculation means, An image processing apparatus, comprising: an identification means for identifying the subject; and an output means for outputting a result of identification by the identification means.