JPH08201022A - Three-dimensional position detecting method and three-dimensional position detecting device - Google Patents

Abstract

PURPOSE: To accurately detect a focusing position even without providing an automatic focusing mechanism by performing a first-order differential filtering to an object image for binary-coding and forming thin lines and then obtaining the total of picture element values on a screen as a feature amount. CONSTITUTION: Light is applied to an object 1 fixed on an x-y-θstage 4 by a lighting device 5 and an image captured by a CCD camera 2 is processed by an image processing device 6 to extract feature amount. Namely, the device 6 performs a first-order differential filtering to the image to extract a contour part and then binaries the image to exclude a prominent and surplus part by the filter. Since the total of the picture element values in an image surface is the feature amount, the size of the image of the object whose image is picked up is affected by the location of an image pick-up position so that the binarized image for reducing the influence is made thin. Then, the total of all picture element values on the screen is obtained as the feature amount. Then, the position of the camera 2 and the state of the stage 4 are controlled by the feature amount, thus detecting the focusing position with less error.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a three-dimensional position detecting method and a three-dimensional position detecting apparatus, and more particularly, to a focus between an image pickup target surface and an image pickup means when picking up an image of an object in image sensing, image recognition or the like. The present invention can be applied to a three-dimensional position detecting method that detects a shift, corrects and measures the distance. Also,
INDUSTRIAL APPLICABILITY The present invention can be applied to, for example, the positioning of an arm of a robot equipped with an autofocus mechanism and a camera, and also detects an angle around an axis of a cylindrical object having a D cut inside and a D cut portion. It can be applied to depth detection and the like. Furthermore, the present invention particularly relates to a three-dimensional position detection method capable of detecting the in-focus position with high accuracy without setting an autofocus mechanism or a distance measuring sensor outside.

[0002]

2. Description of the Related Art In recent years, in a three-dimensional position detecting method in image processing, problems due to autofocus and distance measurement have become quite important. The former method of solving the problem by autofocus is provided with a light projecting means and a light receiving means, and the reflected light emitted from the light projecting means and reflected from the object to be irradiated is received by the light receiving means. Examples thereof include one that adjusts the focus based on the angle, one that divides the captured image into a plurality by a lens group, and adjusts the focus according to their image forming positions.

As a method for solving the latter problem by distance measurement, there are a method using a plurality of cameras, a stereo method using images from a plurality of directions, and a method using a distance sensor. Further, as a method for detecting the focus position in the algorithm of the image processing, for example, a method for detecting a high frequency component by FFT can be cited. Now, by utilizing the fact that the sharpness of the edge part when the image is blurred decreases, it is divided into the required number of pixels or two
An automatic focus control method for a camera that takes the next difference and detects it as the in-focus position when the total sum becomes maximum is disclosed in, for example, Japanese Patent Publication No. 6-16134.

In this conventional camera automatic focus control method, the image clarity determination means for automatically adjusting and focusing the distance between the lens of the camera and the image forming surface,
All or part of the formed image is divided into the required number of pixels, and the absolute value of the difference in brightness or hue between adjacent pixels is extracted to maximize the absolute value of the difference between these adjacent absolute values. As described above, the distance between the lens and the image plane or the focus adjusting means is controlled.

Therefore, since the focus adjustment can be automatically performed while converting the sharpness of the image itself into an electric signal and detecting the electric signal, focus adjustment excellent in responsiveness and suitable for the image, particularly This has an advantage that automatic focus adjustment suitable for an SSTV camera can be performed.

[0006]

However, in the above-mentioned prior arts and the like, when an autofocus mechanism, a distance measuring sensor, etc. are provided, not only the overall system becomes more complicated, but also those autofocus mechanisms. There is a problem in terms of detection accuracy because an offset occurs between the captured image and the distance information obtained by a mechanism such as a distance measuring sensor.

Further, in the above-mentioned conventional example, since the sum of the pixel values in the processed image plane is used as the feature amount, there is a problem that the error is large and the repeat accuracy cannot be obtained. Also,
Even in the actual image situation, changes in shading occur in the contour part and inside of the detection target, so there is a high possibility of erroneous detection only by obtaining the sum of pixel values in the processed image plane, and there is a problem in detection accuracy. was there.

Therefore, an object of the present invention is to provide a three-dimensional position detecting method capable of detecting the in-focus position with high accuracy without setting an autofocus mechanism or a distance measuring sensor outside. .

[0009]

According to the first aspect of the present invention,
A step of capturing an image of an object and capturing an image of the object; a step of applying a first derivative filter to the captured image; and a step of binarizing the image subjected to the first derivative filter; The method is characterized by including a step of thinning the binarized image and a step of subsequently obtaining a total sum of all pixel values on the screen as an extremely small amount.

According to a second aspect of the present invention, in the above-mentioned first aspect of the present invention, a Laplacian filter of four neighborhoods is applied to the captured image, and a difference between the highest pixel value and the lowest pixel value of the image subjected to the Laplacian filter is calculated. When the image is detected and focused, the difference between the highest pixel value and the lowest pixel value of the detected image has a peak at the in-focus position, so that the in-focus position is detected.

According to a third aspect of the invention, in the second aspect of the invention, only the focused portion in the captured image is extracted as a binary image. A fourth aspect of the present invention is characterized in that, in the third aspect of the invention, an area of the binary image of the focused portion is obtained, and the image of the binary image area is extracted from the captured image. It is a thing.

The invention according to claim 5 is the same as claims 1 and 2 above.
In the invention described above, when detecting an object having a step,
It is characterized in that the distance of the step is detected based on the change of the in-focus position extracted with the movement of the imaging system. In the invention according to claim 6, in the invention according to any one of claims 1 to 5, when a step is provided only inside the object,
It is characterized in that the position of the focused portion inside the object is detected.

According to a seventh aspect of the invention, in the first to sixth aspects of the invention, when the image pickup position is deviated from the in-focus position, the brightness of the entire image is compared with a preset brightness, and the comparison is made. Based on the result, it is characterized in that whether the image is deviated in either of the perspective direction is corrected, and the position of the image pickup system is corrected to a distance within a predetermined range. According to an eighth aspect of the present invention, an image pickup unit that picks up an image of an object and captures an image of the object, an image pickup unit moving unit that moves the image pickup unit in a vertical vertical direction, and a stage that fixes the object are horizontally and vertically above and below the horizontal plane. A stage moving means for moving, a rotating means for rotating the stage in a horizontal plane, a light irradiating means for irradiating the object with light, an image processing device for processing an image captured by the image capturing means, the image capturing means, and the image capturing means. A three-dimensional position detection device having a control device for moving a stage, wherein a first-order differential filter processing means for applying a first-order differential filter to the captured image and the image subjected to the first-order differential filter are binarized. And a thinning processing unit for thinning the binarized image, and a feature amount calculating unit for obtaining a sum of all pixel values on the screen as a feature amount. Than it is.

According to a ninth aspect of the present invention, in the above-mentioned eighth aspect of the present invention, the Laplacian filter processing means for applying a Laplacian filter in the vicinity of four to the captured image, and the highest pixel value and the lowest pixel value of the image subjected to the Laplacian filter. A focus position detecting means for detecting the focus position by detecting the difference between the pixel values and, when in focus, detects the difference between the highest pixel value and the lowest pixel value of the detected image at the focus position. It is characterized by having.

The invention described in claim 10 is characterized in that, in the invention described in claim 9, it has a binary image extraction means for extracting only a focused portion in the captured image as a binary image. Is. According to an eleventh aspect of the present invention, in the above tenth aspect of the present invention, a binary image area calculating means for obtaining an area of the binary image in the focused portion,
And an image extracting means for extracting an image of the binary image area obtained from the captured image.

The invention according to claim 12 is the above-mentioned claim 8,
In the invention described in Item 9, when detecting an object having a step, there is provided distance detection means for detecting a distance of the step based on a change in a focus position extracted with the movement of the imaging system. It is a thing. According to a thirteenth aspect of the present invention, in the above-described eighth to twelfth aspects, when a step is formed only inside the object, a position detecting means for detecting the position of the focused portion inside the object is provided. It is what

According to a fourteenth aspect of the present invention, in the above-described eighth to thirteenth aspects, when the image pickup position is deviated from the in-focus position, the comparison means compares the brightness of the entire image with a preset brightness. And a position correcting means for correcting the position of the image pickup system up to a distance within a predetermined range by specifying whether the image is deviated in the perspective direction based on the comparison result.

[0018]

According to the first aspect of the present invention, an image of an object is picked up, an image of the object is captured, and then, in order to extract the sharpness of the contour, a first derivative filter is applied to the captured image, and then the contour is sharpened. The image subjected to the first-order differential filter is binarized in order to make the difference outstanding, and then the binarized image is thinned in order to reduce the influence of the area related to the perspective of the imaging position. The sum of all pixel values is obtained as a feature amount.

For this reason, when performing image processing, a primary differential filter is applied to the image so that the contour of the image becomes clear at the in-focus position without providing an external autofocus mechanism or a distance measuring sensor. Further, by binarizing the image, it is possible to detect the in-focus position where the contour of the image is clear in a series of algorithms for image processing. Moreover,
At that time, by thinning the image, the influence of the increase or decrease of the area of the image with respect to the perspective of the imaging position, that is, the influence of the change of the feature amount due to the shift amount of the focus from the change of the area of the object in the image is reduced. Then, the focus position can be detected so that the error is small.

According to the second aspect of the present invention, in order to enhance the contrast of the image, a Laplacian filter in the vicinity of 4 is applied to the captured image, and the difference between the highest pixel value and the lowest pixel value of the image subjected to the Laplacian filter is detected. Then, when the image is in focus, the difference between the highest pixel value and the lowest pixel value of the detected image has a peak at the in-focus position, so that the in-focus position is detected.

Therefore, when the image processing is performed, the Laplacian filters in the vicinity of 4 are applied to the image, so that the contrast of the image can be made sharp at the in-focus position. Moreover, the difference between the highest pixel value and the lowest pixel value in the Laplacian filtered image plane is used as the feature amount, and the feature amount gives a peak at the in-focus position, whereby the in-focus position can be detected. Therefore, it is possible to detect the in-focus position with higher accuracy at higher speed.

According to the third aspect of the invention, only the focused portion in the captured image is extracted as a binary image. Therefore, when performing image processing, by extracting only the focused portion in the captured image as a binary image, only the image of the attention portion of the three-dimensional object can be extracted as a binary image.

According to the fourth aspect of the invention, the focusing point portion 2 is provided.
The area of the value image is obtained, and the image of the binary image area is extracted from the captured image. Therefore, when the image processing is performed, the area of the binary image of the focused area is specified, and the image of the focused area of the attention area of the three-dimensional object is extracted by extracting the binary image area portion from the captured image. Only the grayscale image can be extracted.

According to the fifth aspect of the invention, when an object having a step is detected, the distance of the step is detected based on the change in the in-focus position extracted with the movement of the image pickup system. Therefore, when performing image processing, when detecting an object having a step, by detecting the step distance based on the change in the in-focus position extracted with the movement of the imaging system, a distance measuring sensor or the like can be externally provided. It is possible to measure the distance such as the height of a step without providing the step.

According to the sixth aspect of the present invention, when there is no step outside the object but only inside the object, the position of the focused portion inside the object is detected. Therefore, when performing image processing, even if there is no step outside the object and there is a step only inside the object, such as a smooth curved surface on the outside, due to the characteristics of the step inside the object, the direction and depth of the step Can be detected in a non-contact manner, and the position and orientation of the entire object can be detected.

Next, when performing image processing, the method according to any one of claims 1 to 6
When the in-focus position is detected by the method described in (1), if it is significantly deviated from the in-focus position, it is possible to specify whether it is deviated to near or near because the change in the feature amount is small if the imaging position moves only a small distance. Can not. Therefore, in the invention according to claim 7, when the image pickup position is deviated from the in-focus position, the brightness of the entire image is compared with a preset brightness, and the image is deviated in either direction of the perspective based on the comparison result. It is configured to correct the position of the imaging system up to a predetermined range of distance.

Therefore, by comparing the brightness in the image plane with the preset brightness, it is specified which direction the image is deviated from, the distance or the near direction, and the image pickup position is quickly corrected up to a certain distance to obtain the in-focus point. Even when the position is considerably deviated from the position, it is possible to specify in which direction, the deviating direction, the position is deviated, and move the image pickup unit to the in-focus position faster. According to the invention described in claim 8, in order to extract the sharpness of the contour, the image captured by the image pickup means is subjected to the first-order differential filter by the first-order differential filter processing means to make the sharpness of the contour stand out. The image subjected to the secondary differential filter is binarized by the binarization processing means, and then the binarized image is thinned by the thinning processing means to reduce the influence of the area relating to the perspective of the imaging position, and then the screen is displayed. The sum of all the above pixel values is obtained as a feature amount by the feature amount calculation means.

For this reason, when performing image processing, an image primary differential filter is applied so that the contour of the image becomes clear at the in-focus position without providing an external autofocus mechanism or distance measuring sensor. By binarizing the image,
The focus position where the contour of the image is clear can be detected by a series of algorithms for image processing. Moreover, at that time, by thinning the image, the influence of the increase or decrease of the image area with respect to the perspective of the imaging position, that is, the influence of the change of the feature amount due to the shift amount of the focus from the change of the area of the object in the image Can be reduced, and the focus position can be detected with less error.

In order to enhance the contrast of the image, the Laplacian filter processing means applies Laplacian filters in the four neighborhoods to the captured image to obtain the highest pixel value and the lowest pixel value of the image subjected to the Laplacian filter. When the difference between the values is detected and the image is in focus, the difference between the highest pixel value and the lowest pixel value of the detected image has a peak at the in-focus position so that the in-focus position is detected by the in-focus position detecting means. Configure.

Therefore, when the image processing is performed, the Laplacian filters in the vicinity of 4 are applied to the image, so that the contrast of the image can be made clear at the in-focus position. Moreover, the difference between the highest pixel value and the lowest pixel value in the image plane that has been subjected to the Laplacian filter is used as the feature amount, and the feature amount gives a peak at the in-focus position, whereby the in-focus position can be detected. Therefore, it is possible to detect the in-focus position with higher accuracy at higher speed.

According to the tenth aspect of the invention, only the in-focus portion in the captured image is extracted as a binary image by the binary image extracting means. Therefore, when performing image processing, by extracting only the focused portion in the captured image as a binary image, only the image of the attention portion of the three-dimensional object can be extracted as a binary image.

In the eleventh aspect of the present invention, the area of the binary image of the focused portion is obtained by the binary image area calculating means, and the image of the binary image area is extracted from the captured image by the image extracting means. To do. Therefore, when the image processing is performed, the area of the binary image of the focused area is specified, and the image of the focused area of the attention area of the three-dimensional object is extracted by extracting the binary image area portion from the captured image. Only the grayscale image can be extracted.

According to the twelfth aspect of the present invention, when an object having a step is detected, the distance of the step is detected by the distance detecting means based on the change in the in-focus position extracted with the movement of the image pickup system. Configure. Therefore, when performing image processing, when detecting an object having a step, by detecting the step distance based on the change in the in-focus position extracted with the movement of the imaging system, a distance measuring sensor or the like can be externally provided. It is possible to measure the distance such as the height of a step without providing the step.

According to the thirteenth aspect of the invention, when there is no step outside the object but only inside the object, the position detecting means detects the position of the in-focus portion inside the object. Therefore, when performing image processing, even if there is no step outside the object and there is a step only inside the object, such as a smooth curved surface on the outside, due to the characteristics of the step inside the object, the direction and depth of the step Can be detected in a non-contact manner, and the position and orientation of the entire object can be detected.

Next, when performing image processing, the method according to any one of claims 8 to 1
When the in-focus position is detected by the method of No. 3, if it deviates considerably from the in-focus position, the change in the feature amount is small if the imaging position moves only a small distance. I can't. Therefore, in the invention according to claim 14, when the image pickup position is deviated from the in-focus position, the comparison means compares the brightness of the entire image with the preset brightness, and based on the result of the comparison, which of the perspective direction is used. It is configured so that the position of the image pickup system is corrected by the position correction means to a distance within a predetermined range.

Therefore, by comparing the brightness in the image plane with a preset brightness, it is specified which direction the image is deviated, that is, the direction in which the image is deviated, and the image pickup position is quickly corrected up to a certain distance. Even when the position is considerably deviated from the position, it is possible to specify in which direction, the deviating direction, the position is deviated, and move the image pickup unit to the in-focus position faster.

[0037]

Embodiments of the present invention will be described below with reference to the drawings. (Embodiment 1) FIG. 1 is a diagram showing the configuration of a three-dimensional position detecting apparatus according to Embodiment 1 of the present invention. In this embodiment, the object 1
The CCD camera 2 that captures the image of is fixed to the camera holder 3 that moves vertically and vertically, and the object 1 is moved horizontally and vertically in the horizontal plane, and also fixed to the xy stage 6 that rotates. The object 1 is irradiated with light by the illumination device 5, and the captured image is processed by the image processing device 6
Then, a desired amount of data is extracted by performing a desired process. The Z-axis control device 7 and the xy- [theta] control device 8 control the position of the CCD camera 2 and the state of the xy- [theta] stage 4 based on these characteristic quantities.

The image processing apparatus 6 of the present embodiment uses a primary differential filter processing circuit for applying a primary differential filter to the captured image and an image for which a primary differential filter is applied in order to emphasize the sharpness of the contour. A binarization processing circuit for binarizing, a thinning processing circuit for thinning a binarized image in order to reduce the influence of the area on the perspective of the imaging position, and a sum of all pixel values on the screen as a feature amount. And a feature amount calculation circuit for obtaining.

Here, the first-order differential filter processing circuit uses, for example, an 8-neighbor filter as shown in FIG. 2, and a first-order differential filter Prewitt as shown in the following equation (1).
Is set and applied to the image to extract the contour portion.

[0040]

[Equation 1]

Although the change of the pixel value is smaller than that of the contour portion, the binarization processing circuit binarizes the image to which the primary differential filter is applied in order to remove the extra portion which is conspicuous by the filter. To do. Since the sum of pixel values in the image plane is used as the feature amount, the size of the image of the object that is imaged affects the distance of the imaging position. Therefore, in order to reduce the influence, the thinning processing circuit thins the binarized image. Then, the feature amount calculation circuit obtains the sum of all pixel values on the screen as the feature amount. Here, the total of the pixel values on the screen is counted and the relationship with the movement in the vertical direction is shown in FIG.

As described above, in this embodiment (claims 1 and 8), in order to extract the sharpness of the contour, the image captured by the CCD camera 2 is processed by the first-order differential filter processing circuit of the image processing device 6 The image subjected to the secondary differential filter and then the primary differential filter in order to emphasize the sharpness of the contour is binarized by the binarization processing circuit of the image processing device 6, and then the area relating to the perspective of the imaging position. In order to reduce the influence of the above, the binarized image is thinned by the thinning processing circuit of the image processing device 6, and then the sum of all pixel values on the screen is used as the feature amount by the feature amount calculation circuit of the image processing device 6. It is configured to ask.

Therefore, when performing image processing, a primary differential filter is applied to the image so that the outline of the image becomes clear at the in-focus position without providing an external autofocus mechanism or a distance measuring sensor. Further, by binarizing the image, it is possible to detect the in-focus position where the contour of the image is clear in a series of algorithms for image processing. Moreover,
At that time, by thinning the image, the influence of the increase or decrease of the area of the image with respect to the perspective of the imaging position, that is, the influence of the change of the feature amount due to the shift amount of the focus from the change of the area of the object in the image is reduced. Then, the in-focus position can be detected with a small error. (Embodiment 2) This embodiment can be applied to the three-dimensional position detecting device of FIG.

The image processing device 6 of the present embodiment includes a Laplacian filter processing circuit for applying a Laplacian filter in the vicinity of 4 to an image captured in order to enhance the contrast of the image, and the maximum pixel value and the minimum pixel value of the image subjected to the Laplacian filter. A focus point position detection circuit that detects the focus point position by detecting the difference between the pixel values, and when the focus is achieved, the difference between the highest pixel value and the lowest pixel value of the detected image has a peak at the focus point position. Have.

Here, FIG. 4 shows a case in which, for example, a C ring as a black object is installed on a stage with a white background, and the in-focus position is detected for the image pickup. The Laplacian filter processing circuit uses the filter of FIG. 2 to set a quadratic differential filter Laplacian shown in the following expression (2), and applies it to the image to emphasize the contrast of the image.

[0046]

[Equation 2]

Then, the in-focus position detection circuit uses the difference between the highest pixel value and the lowest pixel value in the filtered image plane as a feature amount, and as shown in FIG. 5, it has a peak at the in-focus position. The focus position is detected. It should be noted that even if a filter other than this filter is used, it cannot be detected, but if this filter is not used, the maximum pixel value may soon exceed the image gradation and become saturated, which means that It is not preferable in terms of detecting the focal position.

As described above, in the present embodiment (claims 2 and 9), the Laplacian filter processing circuit applies the Laplacian filter in the vicinity of 4 to the captured image and the Laplacian filter in order to enhance the contrast of the image. When the difference between the highest pixel value and the lowest pixel value of the image is detected and the focus is achieved, the difference between the highest pixel value and the lowest pixel value of the detected image becomes the peak at the in-focus position, so that the in-focus position is brought into focus. The position detection circuit is configured to detect.

Therefore, when the image processing is performed, the Laplacian filters in the vicinity of 4 are applied to the image, so that the contrast of the image can be made sharp at the in-focus position. Moreover, the difference between the highest pixel value and the lowest pixel value in the image plane that has been subjected to the Laplacian filter is used as the feature amount, and the feature amount gives a peak at the in-focus position, whereby the in-focus position can be detected. Therefore, it is possible to detect the in-focus position with higher accuracy at higher speed. (Embodiment 3) This embodiment can be applied to the three-dimensional position detecting device of FIG.

The image processing apparatus 6 of this embodiment has a binary image extraction circuit for extracting only the focused portion in the captured image as a binary image. Here, for example, FIGS. 6 and 7 show a case in which only the image of the D-cut 11 of the cylindrical object having the D-cut 11 inside and having no step outside is taken out. FIG. 6 shows a state of the D-cut 11 applied inside the object, and FIG. 7 shows a state of an image from above the object. In FIGS. 6 and 7, 12 is a cylindrical mouth.

FIG. 8 shows a schematic diagram of an image subjected to Laplacian when the D-cut 11 part is focused. FIG. 9 shows a schematic diagram of an image in which the focus position where a relatively high frequency component exists remains, and which is binarized after pre-processing of expansion and compression. As described above, in this embodiment (claims 3 and 10), only the in-focus portion in the captured image is converted into a binary image by the binary image extraction circuit.
It is configured to be extracted as a value image.

Therefore, when performing image processing, by extracting only the focused portion in the captured image as a binary image, only the image of the attention portion of the three-dimensional object can be extracted as a binary image. . (Embodiment 4) This embodiment can be applied to the three-dimensional position detecting device of FIG.

The image processing apparatus 6 of this embodiment comprises a binary image area calculation circuit for obtaining an area of a binary image in a focused area and an image extraction circuit for extracting an image of a binary image area obtained from a captured image. Have and. Here, for example, the coordinates of each pixel of the portion of the black image of FIG. 9 described in the third embodiment are obtained, and only the portions corresponding to those regions in the grayscale image captured first are left, and the others are black or white. As the image, only the focused portion is extracted as the grayscale image.

Thus, the present embodiment (claims 4 and 11)
Then, the binary image area of the focused portion is obtained by the binary image area calculation circuit, and the image of the binary image area is extracted by the image extraction circuit from the captured image.
Therefore, when the image processing is performed, the area of the binary image of the focused area is specified, and the image of the focused area of the attention area of the three-dimensional object is extracted by extracting the binary image area portion from the captured image. Only the grayscale image can be extracted. (Embodiment 5) This embodiment can be applied to the three-dimensional position detecting device of FIG.

The image processing device 6 of this embodiment detects a distance such as a step based on a change in the in-focus position extracted with the movement of the image pickup system when detecting an object having a step. Have. Here, for example, as shown in the case of detecting the distance between the D-cut 11 portion and the cylindrical mouth portion 12 of FIGS. 6 and 7 described in the second embodiment, the D-cut 11 portion of the D-cut 11 portion is detected by the method of the second embodiment. The position is detected, the CCD camera 2 is further moved upward, and the cylindrical mouth portion 12 is detected again by the method of the second embodiment described above. CC at that time
FIG. 10 shows the relationship between the movement of the D camera 2 and the change in the characteristic amount.

Thus, the present embodiment (claims 5 and 12).
Then, when detecting an object having a step, the distance detection circuit detects the distance of the step based on the change in the in-focus position extracted with the movement of the imaging system.
Therefore, when performing image processing, when detecting an object having a step, by detecting the step distance based on the change in the in-focus position extracted with the movement of the imaging system, a distance measuring sensor or the like can be externally provided. It is possible to measure the distance such as the height of a step without providing the step. (Embodiment 6) This embodiment can be applied to the three-dimensional position detecting device of FIG.

The image processing apparatus 6 of the present embodiment has a position detection circuit for detecting the position of the in-focus portion inside the object when the object has a step only inside the object. Here, for example, regarding the object of FIGS. 6 and 7 described in the second embodiment, as shown in the case of detecting the direction of the D cut 11, the image of the D cut 11 is extracted as a binary image by the third embodiment described above,
Find its center of gravity. In addition, the outline of the cylindrical mouth portion 12 is regarded as a circle, its center coordinates are obtained, and the D cut 1 is made from the center coordinates.
The direction of 1 is detected. A schematic diagram of this detected image is shown in FIG.
Shown in

Thus, the present embodiment (claims 6 and 13).
Then, when there is no step outside the object but only inside the object, the position detection circuit detects the position of the focused portion inside the object. Therefore, when performing image processing, even if there is no step outside the object and there is a step only inside the object, such as a smooth curved surface on the outside, due to the characteristics of the step inside the object, the direction and depth of the step Can be detected in a non-contact manner, and the position and orientation of the entire object can be detected. (Embodiment 7) This embodiment can be applied to the three-dimensional position detecting device of FIG.

The image processing apparatus 6 of the present embodiment, when the image pickup position is deviated from the in-focus position, compares with the comparison circuit for comparing the brightness of the entire image with the preset brightness, and the perspective based on the comparison result. It has a position correction circuit that identifies in which direction the deviation is made and corrects the position of the imaging system up to a distance within a predetermined range. Here, for example, in the above-described second embodiment, when the object of FIG. 4 is detected, as can be seen from FIG. 5, the change in the feature amount is remarkable at the position close to the in-focus position, but it is significantly deviated from the in-focus position. On the other hand, there is no change in the feature quantity, so it is difficult to predict whether the distance is far or near. Therefore, as shown in FIG. 12, it is predicted from the brightness of the image which side of the image is greatly shifted, the CCD camera 2 is quickly moved to near the in-focus position.

As described above, this embodiment (claims 7 and 14)
Then, when the imaging position is deviated from the in-focus position, the brightness of the entire image is compared with the preset brightness by the comparison circuit, and based on the comparison result, it is specified in which direction the distance is deviated, The position correction circuit corrects the position of the imaging system up to a distance within a predetermined range.

For this reason, by comparing the brightness within the image plane with the preset brightness, it is specified which of the perspective and the direction is deviated, and the imaging position is quickly corrected up to a certain distance to obtain the in-focus point. Even if it deviates considerably from the position, identify in which direction it is deviating, the CCD
The camera 2 can be moved faster to the in-focus position.

[0062]

According to the present invention, there is an effect that the in-focus position can be detected with high accuracy without setting an autofocus mechanism or a distance measuring sensor outside.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of a three-dimensional position detection device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing an eight-neighborhood filter according to a first embodiment of the present invention.

FIG. 3 is a diagram showing a relationship between a total of pixel values on a screen and a movement in a vertical up and down direction according to the first embodiment of the present invention.

FIG. 4 is a diagram showing a case where a C ring as a black object according to a second embodiment of the present invention is installed on a stage with a white background, and a focus position is detected for the imaging thereof.

FIG. 5 is a diagram showing a state in which a peak occurs at a focus position of Example 2 according to the present invention.

FIG. 6 is a diagram showing a D-cut state applied to the inside of a cylindrical object having no step on the outside according to Example 3 of the present invention.

FIG. 7 is a diagram showing a state of an image from above of a cylindrical object having no step on the outside according to the third embodiment of the present invention.

FIG. 8 is a diagram showing an image subjected to Laplacian when a D-cut portion of Example 3 according to the present invention is focused.

FIG. 9 is a diagram showing an image obtained by performing binarization after performing a pre-expansion / compression pre-processing with a focus position where a relatively high-frequency component exists according to the third embodiment of the present invention.

FIG. 10 is a diagram showing a relationship between movement of a CCD camera and change in feature quantity according to a fifth embodiment of the present invention.

FIG. 11 is a diagram showing an image in which the direction of D-cut is detected from the center coordinates where the contour of the cylindrical mouth portion of the sixth embodiment according to the present invention is regarded as a circle.

FIG. 12 is a diagram showing a sum of pixel values with respect to defocus according to a seventh embodiment of the present invention.

[Explanation of symbols]

 1 Object 2 CCD Camera 3 Camera Holder 4 xy-θ Stage 5 Illumination Device 6 Image Processing Device 7 Z-Axis Control Device 8 xy-θ Control Device 11 D Cut 12 Cylinder Mouth

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G06T 5/20 H04N 5/232 A G06F 15/68 400 A

Claims (14)

[Claims] 1. A step of capturing an image of an object and capturing an image of the object, a step of applying a primary differential filter to the captured image, and then binarizing the image subjected to the primary differential filter. A three-dimensional position detecting apparatus, comprising: a step of performing a thinning process on the binarized image; and a step of determining a total sum of all pixel values on the screen as an extremely small amount. 2. A Laplacian filter in the vicinity of 4 is applied to the captured image, the difference between the highest pixel value and the lowest pixel value of the image subjected to the Laplacian filter is detected, and when the image is in focus, the highest image value of the detected image is detected. The three-dimensional position detecting method according to claim 1, wherein the in-focus position is detected when the difference between the pixel value and the minimum pixel value reaches a peak at the in-focus position. 3. The three-dimensional position detecting method according to claim 2, wherein only the focused portion in the captured image is extracted as a binary image. 4. An area of the binary image of the focused portion is obtained,
4. The three-dimensional position detecting method according to claim 3, wherein an image of the binary image area is extracted from the captured image. 5. When detecting an object having a step, the distance of the step is detected based on a change in the focus position extracted with the movement of the imaging system. 3D position detection method. 6. The three-dimensional position detecting method according to claim 1, wherein when a step is provided only inside the object, the position of the focused portion inside the object is detected. 7. When the image pickup position is deviated from the in-focus position, the brightness of the entire image is compared with a preset brightness, and it is specified whether the image is deviated in the near or far direction based on the comparison result. 7. The three-dimensional position detecting method according to claim 1, wherein the position of the imaging system is corrected to a distance within a predetermined range. 8. An image pickup means for picking up an image of an object and capturing an image of the object, an image pickup means moving means for moving the image pickup means in a vertical vertical direction, and a stage for moving a stage for fixing the object vertically and horizontally in a horizontal plane. Moving means, rotating means for rotating the stage in a horizontal plane, light irradiating means for irradiating the object with light, an image processing device for processing an image captured by the image capturing means, and moving the image capturing means and the stage. A three-dimensional position detecting device having a control device for
Primary differential filter processing means for applying a primary differential filter to the captured image, binarization processing means for binarizing the image subjected to the primary differential filter, and thinning the binarized image. A three-dimensional position detecting device comprising: a thinning processing means and a feature amount calculating means for obtaining a sum of all pixel values on a screen as a feature amount. 9. A Laplacian filter processing means for applying a Laplacian filter in the vicinity of 4 to the captured image, and a difference between the highest pixel value and the lowest pixel value of the image subjected to the Laplacian filter is detected, and when the image is in focus, it is detected. 9. The three-dimensional position according to claim 8, further comprising: focus position detecting means for detecting the focus position by the peak of the difference between the highest pixel value and the lowest pixel value of the image at the focus position. Detection device. 10. A three-dimensional position detecting apparatus according to claim 9, further comprising a binary image extracting means for extracting only a focused portion in the captured image as a binary image. 11. A binary image area calculating means for obtaining an area of a binary image in the focused portion, and an image extracting means for extracting an image of the binary image area obtained from the captured image. The three-dimensional position detection device according to claim 10, which is characterized in that. 12. When detecting an object having a step, based on a change of a focus position extracted with a movement of an imaging system,
10. The three-dimensional position detecting device according to claim 8, further comprising a distance detecting means for detecting a distance of the step. 13. A three-dimensional position detecting apparatus according to claim 8, further comprising position detecting means for detecting a position of a focused portion inside the object when there is a step only inside the object. . 14. If the image pickup position is deviated from the in-focus position, whether the image is deviated in either of the near and far directions based on the result of the comparison means for comparing the brightness of the entire image with the preset brightness. 14. The three-dimensional position detecting device according to claim 8, further comprising: a position correction unit that specifies the position of the image pickup system and corrects the position of the image pickup system up to a predetermined range of distance.