JPS59112211A - Method and device for determining space coordinate of workpiece - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

TECHNICAL FIELD OF THE INVENTION The present invention relates to a fast scanning method and apparatus for determining the position of a sensed electrical pulse, and more specifically, the spatial coordinates of a workpiece within a field of view. The present invention relates to improvements, particularly to high speed scanning methods and apparatus that can provide faster and more accurate spatial coordinates of a workpiece within a field of view than previously known.

[Technical background of the invention]

BACKGROUND OF THE INVENTION Devices are known in the art that are capable of viewing a workpiece, determining its position, and providing instructions to equipment acting on the workpiece. However, these conventional devices suffer from slow response times and poor accuracy due to specific limitations in the cameras and laser scanning used; for example, as disclosed in U.S. Pat. No. 4,188,544. The device was designed for use in the wood industry, and uses a light source to illuminate a log as a workpiece with a planar beam of light, and a vidicon camera to detect the intersection line between the log surface and the planar beam of light. It is used to scan. This publication also discloses a calibration technique for correcting the signal from the image pickup tube. According to this technology, calibration information obtained by comparing a scanning signal from a known three-dimensional reference object placed in the field of view of the camera with previously obtained spatial coordinates of that reference object is stored in a memory unit. store. The actual measurement is performed by comparing the scan signal of the workpiece with the calibration information obtained by a calibration scan. In this prior art system,
Only the spatial coordinates stored in memory are used for calibration calculations. For this reason, in the conventional apparatus, there are limits to both the number and precision of the spatial coordinate data that can be stored, and the resolution thereof, and there is also a limit to the ability to accurately determine the corresponding spatial coordinates during caliper plane scanning.

U.S. Patent Nos. 3,895,870 and 3,796
, No. 492 discloses another system in which coherent light is irradiated onto the surface of a workpiece and the image is received by a vidicon image pickup tube camera. What is disclosed in these patents is a combination of a linear beam of the body and a pair of image receiving parts,
Alternatively, the combination of a pair of linear beams and an image receiving section determines the distance between the object being irradiated with the beam and the camera of the image receiving section.In other words, it is simply a distance meter. Only.

US Pat. No. 4,111,557 discloses a method for optically determining and comparing the shape and position of objects.

In this method as well, an object is imaged by a photoelectric receiver, converted into electrical pulses, and stored as spatial coordinate data after AD conversion. Data about the actual object after it has moved or deformed is compared with previously stored data to determine changes in the object's position and shape.

However, this invention merely detects changes in position and shape, and cannot even attempt to determine spatial coordinates.

U.S. Pat. No. 4,146,926 also discloses a method for detecting the surface of an object and determining its position with respect to a reference point by a combination of a linear beam and a pair of reflected light focusing means. ing.

U.S. Patent Nos. 4,115,806 and 3.980
．． Although the disclosure of No. 812 relates to a method for analyzing an image converted into an electrical signal, it is considered to have very little relevance, unlike the method according to the present invention.

US Pat. No. 4,126,395 discloses a method for determining the spatial position of an identification mark on a mirror surface as the position of each point.

U.S. Patent No. 3,459,246, U.S. Patent No. 3,736,9
No. 68, No. 3.7137.700, and No. 3,
No. 963,938 discloses another prior art system using a non-coherent light source and a phototube sensor, and U.S. Pat. A system for calculating the size of an object using known distances and geometric relationships is disclosed.

U.S. Patent No. 3,187,185, U.S. Patent No. 3.590.2
No. 58 and No. 3,625,618 disclose a system for determining the external surface of an object by observing the object from a direction at an angle of 8v with respect to the illuminating light beam.

A system using a conventional sensor and a video digitizer, such as the one used in the prior art, had a problem with the accuracy of the position determination pressure of the detection pulse. The method of determining the pulse center of gravity in the present invention is a new technique that has not been used in these prior art techniques. The present invention also includes an improved method of calibration techniques for accurately determining the spatial coordinates of an object.

[Purpose of the invention]

The present invention has a general object of providing an apparatus and method for determining the spatial coordinates of a workpiece, and more specifically has the following objects.

A first object of the present invention is to provide an apparatus and method that allows the spatial coordinates of a workpiece placed within a field of view to be determined more accurately and faster than in the prior art.

A second object of the invention is to provide an apparatus and method for determining the overall dimensions and visual details of workpieces with irregular surfaces.

A third object of the invention is to provide an apparatus and method for determining details of visual elements of workpieces with a wide variety of optical properties.

A fourth object of the present invention is to provide an apparatus and method that allows self-calibration without having to provide a jig with a known shape within the field of view and record the spatial coordinates of this jig. .

A fifth object of the present invention is to determine the centroid of a pulse by using relatively inexpensive analog or digital logic circuits, such as adders, without the need for expensive and time-consuming multiplication processes. The objective is to provide a device and method that can.

[Summary of the invention]

The present invention relates to a method and apparatus for determining the position of a pulse center of gravity, and a high-speed scanning method and apparatus for determining the spatial coordinates of an object using this method or pulse integration method. The method according to the invention also includes the step of illuminating the object with a coherent light source. An image of the object including the intersection line between the coherent light and the object surface is formed using a conventionally known Senshita filter and lens in a photoelectric conversion device such as an image orthicon tube. A photoelectric conversion device receives an image of an object at a position separated by a certain angle from a coherent light source. The intersection of the planar beam of coherent light and the object surface generates one reference pulse on each horizontal scan line of the image orthicon tube. The position of each pulse, in particular the geometric mean or center of gravity of that pulse, is determined.

Two pulse position determination methods are disclosed herein.

The first method is generally called the "pulse integration method," in which one signal pulse from a photoelectric conversion device is first sampled and digitized. Subsequently, pulse integration is performed in the following manner. First, each sample value of the sampled pulses is summed. Next, each sample value is doubled.

Then, from the first sample value to the last sample value (
or vice versa), then subtract twice the value of each sample value one after another from the previously determined total value (.By such subtraction, the previously determined total value gradually decreases. However, when this total value exceeds 0, this subtraction operation is stopped.Here, by subtracting twice that value from the total value, the sample whose total value exceeds O is used as the base unit. Record it as a sample.Finally, find the ratio of the value twice the basic sample value and the absolute value of the total value at the time of the subtraction operation,
The centroid position of the pulse is determined by adding (or subtracting) a value corresponding to this ratio to the position of the basic single sample.

The second method is generally referred to as the "center of gravity method" and, like the "pulse integration method", one signal pulse from the photoelectric conversion device is first sampled and digitized. Next, the center of gravity is determined using the following method. First, a first total value, which is the sum of each sample value of the sampled pulses, is determined. Next, the width value for each sample value is determined, but this is determined as the sum of all sample values before (and including) that sample value. Subsequently, a second total value is calculated by summing all the middle 11 values for each sample value. Then, the position of the last sample value of the pulse is recorded, and by subtracting the value corresponding to the ratio of the first sum value and the second sum value from the previously recorded position, the centroid of the pulse, that is, the geometric Determine the average.

In the above two methods, it is preferable to use a digital logic circuit. However, in the center of gravity method, it is also possible to use an analog integrator. The angular and distance relationships between the coherent light source, the workpiece, and the lenses of the photoelectric converter are determined by a unique calibration method described below, and the geometric mean of the reference pulse, i.e., the center of gravity, is determined by this angular and distance relationship. is determined as described above, and the spatial coordinates of the workpiece are thereby determined.

The calibration technique is performed by placing a precisely shaped jig within the field of view of the photovoltaic device. A geometric figure with a unique cross-sectional shape, such as a triangle, is uniquely determined by the jig. The dimensions of this geometry, rather than the spatial coordinates of the jig, are stored in detail in the connected computer. The shape of the geometry is determined by scanning points along the flat surface of the jig, connecting the points with lines, and the intersections of the lines. These intersection points will determine the vertices of the unique geometry on the jig. In this way, the spatial coordinates X, Y, and Z of the workpiece within the field of view of the photoelectric conversion device can be determined by the geometrical figure obtained by referring to the dimensions previously stored for the figure. It becomes possible to calculate accurately. It is preferable that the longitudinal direction of the jig is arranged along the horizontal scanning line of the photoelectric conversion device.

However, it is not necessary to make them completely parallel.

A preferred embodiment of the present invention will be described in detail below with reference to the accompanying drawings. This will make the various features, objects, and advantages of the invention described so far clearer.

[Preferred embodiments of the invention]

A typical example of an apparatus for determining the spatial coordinates of a workpiece, which is an embodiment of the present invention, will be described with reference to the block diagram of FIG. An optical system 12 is connected to a commonly used light source such as a laser 10.
is attached to create a planar beam of coherent light, and the workpiece 14 is irradiated with this. In Fig. 1, a flat plate of wood is used as an example of the workpiece 14, but it goes without saying that the workpiece may be any object whose spatial coordinates need to be determined for precision measurement, inspection, or processing. stomach. Therefore, although the present invention will be described here with reference to the specific example of processing wood, the apparatus according to the present invention may also be used in various industrial control devices, production line inspections, inventory management, medical scanning, etc. It can be used without restrictions. Furthermore, the pulse position determination method described later is a technology that can be widely applied to fields other than wood processing.

An image pickup tube, such as an image orthicon, vidicon, or ultracon, is used as the photoelectric conversion device 18 and is positioned at a known, predetermined angle with respect to a planar light source to capture an image of the workpiece. This image pickup tube generates multiple scan lines per scan figure. Each scan line has one reference pulse on it. This reference pulse corresponds to the intersection of the scanning line and the planar beam of coherent light that is irradiating the workpiece. When viewed on a video imager such as that shown in FIG. 1, a reference pulse line formed by the reference pulses of each scan line can be seen. The image tube output signal is input to a video digitizer 22, where sampling and digitization including reference pulses are performed for each scan line. Video digitizer 22 uses the digitized information to determine the position of each reference pulse, and in particular the position of the centroid, or geometric mean, of each pulse.

This determined accurate IL position information of each pulse, optical system 1
A jig with a known shape placed within the field of view of 2 and an optical system 1
2 and various known information determined by the calibration techniques described below are integrated by a computer, so that the spatial coordinates of the workpiece are gradually determined. The spatial coordinates thus determined are input into another computer which controls an external device used by the user, such as a lumber cutting machine. Although the present invention does not include such devices used or present by a user, it is well known that the output of the present invention can be easily used to control such devices.

Another embodiment of the present invention is schematically shown in FIG.

As in the previous embodiment, the laser 1o serves as a scanning light source and irradiates the workpiece 14 with coherent light. two separate,
Moreover, the optical system 2 has image receptors separated from each other by a certain distance.
0 captures the workpiece and creates two images of the workpiece. The two images created in the optical system 2o are received by a photoelectric conversion device R18, such as an image orthicon tube, and a pair of reference pulses are created for each scanning line. These reference pulses correspond to the intersections of the scanning line and the planar beam of coherent light that is irradiating the workpiece. When this result is displayed on a video imager as shown in FIG. 2, it becomes a pair of reference pulse lines as shown in the figure. Similar to the previous embodiment, one or more video digitizers 22 sample and digitize each scan line containing two reference pulses. This digitized information is used to very precisely determine the location of the centroid, or geometric mean, of each pulse, and this information, along with known or calculated parameters, is used in Figure 1. The spatial coordinates of the workpiece are determined by a method as will now be described in detail based on the apparatus.

It should be noted that the video imager shown in FIGS. 1 and 2 is not included in the equipment necessary to practice the present invention. Here, the video imager is merely used as an example to show the visual form of the electrical signals input to the video digitizer.

The embodiment shown in FIG. 1 will be explained again. Laser 10 is shown separate and apart from the other elements of the system to indicate that the invention does not require any hardware interface between laser 10 and the other elements. If the most obvious hardware interface were to be used, it would be a laser angle detector used to calculate the X coordinate of the laser as it travels. A laser 1° placed above the workpiece emits a line of coherent light, which is converted into a planar beam by a conventionally known optical system 12 for the laser output light. Although it is possible to use light sources other than lasers in the practice of the present invention, it is preferable to use lasers as the light source to avoid scattering effects of ambient light.

A planar beam of coherent light is directed onto a workpiece 14 whose spatial coordinates are to be determined. The workpiece is placed on a conveyor 16 or the like, a portion of which is shown in the figure. Although in the preferred embodiment the laser 10 is fixed so that it does not move, another embodiment of the invention is to mount the laser so that it can move only laterally to scan the workpiece. Good too.

As for the lens used in the photoelectric conversion device 18, in the device shown in Fig. 1, a normal lens is used which is appropriately selected to ensure that the workpiece can be kept within the field of view, but various lenses may be used depending on the purpose of the work. is selected. Although not shown in the figure, a laser interference filter is provided between this lens and the photoelectric conversion device to cut out ambient light. In an optical system having two image receptors as shown in FIG. 2, lenses for binoculars are appropriately selected to the extent that they can be secured within the processing field.

The photoelectric conversion device shown in FIG. 1 may be a standard closed circuit television camera that repeatedly sweeps from left to right and generates multiple scan lines per scan pattern.

The camera can be of high resolution, low resolution, high sensitivity, or low sensitivity, depending on the requirements of the task of the device. Each scan line contains a reference pulse corresponding to the intersection of that scan line with a planar beam of coherent light that is being applied to the workpiece, producing an output signal as shown in the video imager of FIG. Form.

Video digitizer 22, shown as one block in FIGS. 1 and 2, is shown in more detail in FIG. This video digitizer 22 receives an analog signal that is the output of the photoelectric conversion device 18, and generally speaking, samples and digitizes this analog signal. 1 from the image orthicon tube
After one sweep of the signal has been sampled, the geometric mean, or center of gravity, of the reference pulse is determined by digital logic circuitry within video digitizer 22 in the manner described below. The output of the video digitizer 22 is input to a computer U, which is shown diagrammatically in FIG. 1 as a collection of blocks representing each function. The computer U that receives the output signal of the video digitizer linearizes this signal, that is, mathematically compensates for nonlinear parts such as the laser light, the camera lens, the camera's electrical components, etc. However, such compensation may only be necessary if the camera system used produces nonlinear images that significantly affect the results. With a high-quality camera and laser lens, such linear compensation is necessary, whereas a lower-quality lens would distort accuracy considerably.

The output of video digitizer 22 is then combined with calibration parameters to calculate the spatial coordinates of the workpiece. This calibration parameter is determined by scanning a precisely shaped jig, selecting each scanned point along the flat surface of the jig, and connecting these points to form a line of intersection. Calculated. The intersection points of the intersection lines thus formed are stored in combinations. These intersection points determine the vertices of the jig-specific geometry, such as a triangle, and also determine the dimensions, such as the base and height. By combining the geometry obtained by scanning the jig with the dimensions memorized for that geometry, the calibration parameters are calculated by the method described below and are calculated for each scanned point on the workpiece. The position is accurately determined.

An important advantage and improvement provided by the invention is that the accuracy in determining the position of the reference pulse to be determined is significantly improved by the device according to the invention. With the bandwidth of typical video signals used in television broadcasting and the like, the horizontal resolution is limited to about 300 to 500 pixels. According to the apparatus and method according to the present invention, the potential usable resolution can be increased from about 20,000 pixels to 10,000 pixels.
Extremely high resolution up to about 0.00 pixels or more can be obtained.

As previously mentioned, the centroid, or geometric mean, of the reference pulse produced by the photoelectric conversion device is determined by one of two methods.

One is the pulse integration method, in which the area of a pulse is determined by integration, and a position is found where this area is divided into two equal parts. The other method is the centroid method, which calculates the single integral value and double integral value of sampled values of the pulse. Here, the double integral value is the ordinary integral value, that is, the sum of all sampled values of the pulse, and the double integral value is, as will be explained in detail later,
-It is the sum of all widths determined as multiple integral values. The offset value is determined by dividing the value obtained by double integration by the value obtained by single integration, and the centroid of the pulse is determined by subtracting this offset value from the position of the last sample value of the pulse. . In this embodiment, the threshold value for pulse integration, which is limited by the resolution of the video digitizer, is preset by the computer section of the apparatus. A video digitizer samples a video signal and generates a sample value of a digital signal by amplifying the sampled value several times.

In the first method, the pulse integration method, all amplified samples of the pulse are stored. This is because according to the invention, these sample values are required in the pulse integration process. Both methods involve linear interpolation or, where possible, curved interpolation using polynomials. These interpolations allow us to achieve
The average of the pulses can be determined with as much precision as possible. Although either method can be implemented using analog circuits, bit slice processors, processor algorithms implemented as software, or digital circuits, the method provides the most concrete solution to the current technical problem. would be a method using digital circuits.

Using the block diagram in Figure 3, the video digitizer 2
The first method performed by No. 2, that is, the pulse integration method, will be explained. Video signal is video preamplifier 302
The signal is matched, buffered, and then input to an A/D converter 304 that digitizes the video analog signal.

The signal from the video preamplifier 302 is also passed through a sync detector 306 and a frame detector 308 to provide a sync pulse at the start position of each video scan line and a sync pulse at the start point of the video screen (at the top of the screen), respectively.
This is the frame (blanking) pulse at . These pulses are input to a computer 24, and the scan line number p scanned by the video camera is determined. Synchronous oscillator 310 receives the synchronizing pulse and generates an output signal synchronized with the synchronizing pulse. This output signal can therefore be used as logic timing within the video digitizer 22, and can also be used to sample the entire video signal.

On the other hand, the A/b converter 304 is connected to the video preamplifier 302.
sample the video signal that is buffered in
The sample value is converted into binary data (digital). The video digitizer 22 has two operating modes: (
1) scan mode and (2) search mode. In the scan mode, the digitizer digitizes, refines, and stores each sample value after the sample value is greater than or equal to a set value set by the computer. Sampling is continued in this manner until the sample value becomes equal to or less than the set value, and each sample value is integrated (integrated) and stored. While sampling is occurring, counter 322 counts N pulses from synchronous oscillator 310. The pulses from synchronous oscillator 310 are thus counted within counter 322 as a count value N, which corresponds to the approximate position on the scan line of a given scan figure at which the corresponding specimen was sampled. This rough position N (that is, the rough count value N) becomes a more accurate value by successively obtained finer correction elements. This correction element will be described below.

After all sample values are accumulated, each accumulated sample value is stored in random access memory (RAM) 314. The integration is performed by an adder 316, and the integrated sum Σ of all sample values of /<A/s, ie, the total sample value representative of that pulse, is stored in a register 317. During the scan mode, the sign of each sample value added by adder 316 is inverted by scaler-ais before being added to adder 31 so that the accumulated sum value to be stored in register 317 is negative.
6 is input.

Once all the pulse samples have been added, the system switches to search mode. In search mode, Memo IJ-314
An adder 316 adds twice the value of each sample value stored in the register 317 to the value in the register 317. The scaler 318, which performs sign inversion for each scan,
In the search mode, an operation to double the sample value is performed under the control of the video pulse detector 320. Each time this double value is added to the adder, the sign bit of the contents of register 317 that is the result of the accumulation is examined by search control logic circuit 324. If this sign bit becomes positive (
・The following three values are stored. First, counter 322
The count value N is stored in the N register 326.

Next, a value twice the last sample value A to which the addition was performed, ie, 2A, is stored in the 2A register 328. Finally, the positive number remaining in the register 317, that is, the addition remainder Σr is Σ
r register 330. By substituting these three values into the following equation, we can calculate the geometric mean η of each scanning pulse.
is required.

Σr η-N-) ----- A Also, when switching from scan mode to search mode, register 3
The sum of all sample values of the pulses remaining in number 17, ie, the area of the pulses, is stored in the B register 332.

Considering the pulse integration method in detail, the A/D converter 30
Maximum resolution can be obtained from the native A/D converter used within the video digitizer by setting the limits of the A/D converter to take advantage of the full scale of the digital output signal that can be generated by the video digitizer. can.

Let us summarize the above pulse integration method once again. When the magnitude of the rising portion of the pulse exceeds a set value, the video digitizer 22 samples the output signal from the image pickup tube and digitizes it. These digitized values are individually stored in the RAM 314, and their signs are inverted and added together as a negative total value. Once all sample values representative of a pulse have been individually stored and summed, the individual sample values stored in RAM 314K are sequentially retrieved, multiplied by 2, and added to the negative sum value by adder 316. .

That is, all the sample values representing the pulse are taken one by one in the order in which they were stored starting with the first sample value, or in the reverse order in which they were stored starting with the last sample value, multiplied by 2, and then multiplied by 2.
The multiplied values are successively added to the total value by the adder 316. This operation continues until the total value of the adder 316 becomes a positive value, that is, the sum of the added double sample values is 2Σ
1 continues until the sum of sample values Σ is exceeded.

Once the adder 3160 sum is positive, it is known that the location of the last sample value added is within one sample interval from the centroid or geometric mean of the pulse. In this way, the corresponding counter value remaining in the N counter (
N) before interpolation is used as the reference sample position to determine the geometric mean. If the operation of doubling the sample value and adding 9 is performed from the last sample value of all sample values representing the pulse, the N counter will continue counting by decreasing by l, but the sample value If the operation of doubling and adding is performed from the -@first sample value of all sample values representing the pulse, the N counter is reset to the value corresponding to the first sample value, From there, it is important to increase the count by 1. The interpolation correction of the reference sample position, that is, the value of the N counter, is performed by adding the addition remainder Σr, which is the final sum remaining in the adder 316, and multiplying it by 2 to the sum of the adder 316.
It is calculated using the last sample value A added as A. The amount of interpolation is determined by dividing the addition remainder Σ by 2A, which is twice the final addition sample value. Therefore, the center of gravity of the pulse, ie, the geometric mean η, is determined by the following equation.

Σr η−N±□ A Here, if the double addition is performed in order from the first sample value of the pulse, the interpolated value Σr/2A is added to N, and the double addition is performed in reverse order from the last sample value of the pulse. , the interpolated value Σr/2A is subtracted from N.

The fact that the geometric mean is found by adding or subtracting the interpolated value to N means that Σr is the last number added to the adder 316, that is, the final summed sample value whose sum exceeds the geometric mean. This is obvious if we consider that this corresponds to a proportional allocation value for a double value. In this way,
The exact location of the geometric mean is determined by adding or subtracting the interpolated value Σr/2A to the N counter value indicating the location of the final summed sample value. The digital resolution in this case is Mx2A
' is given by '. Here, M is the number of samples constituting one scanning line, and A' is the maximum number of digital steps allowed for one sample value. For example, when using an 8-bit A/D converter, the maximum number of digital stages allowed for one sample value is 2.
It is 56. Therefore, if an 8-bit '/D converter is used and sampling is performed at a sampling period of 30MH2 (M=-1700), MX2A'-8,70x 10
becomes.

Next, the centroid method performed by the video digitizer 23, which can be used in place of the video digitizer 22 described above, will be described with reference to the block diagram shown in FIG. Similar to the pulse integration method described above, the video signal is input to a video preamplifier 402 where it is matched, buffered, and then input to an 'A converter 404. Also, the signal from this video preamplifier 402 is
Via a sync detector 406 and a frame detector 408, a sync pulse at the start of each video scan line and a frame (blanking) pulse at the start of the video screen (top of the screen), respectively. These pulses are input to the computer 24, and the scan number P scanned by the photoelectric conversion device 18 is determined. Synchronous oscillator 410 receives the synchronizing pulse and generates an output signal synchronized with the synchronizing pulse. This output signal can therefore be used as logic timing within the video digitizer 23, and can also be used to sample the entire video signal.

On the other hand, the A/D converter 404 is connected to the video preamplifier 402.
sample the video signal that is buffered in
The sample value is converted into binary data (digital). 1st
1! Ifg unit 416 sums each sample value after the sample value exceeds a set value set by computer 24 . Sampling continues in this way until the sample value falls below the set value, and each sample value is the M'A, C integral)
be done. This is the first total value. The second total value is obtained by summing each intermediate value obtained each time the first adder 416 adds each sample value, using the second adder 418.

While sampling is occurring, counter 422 counts as N pulses from synchronous oscillator 410.

Each pulse from synchronous oscillator 410, i.e. counter 422
The count value in corresponds to each sample of the video signal. The count value N of counter 422 also indicates the position of the corresponding sample value on a given scan line. The higher the frequency of the synchronous oscillator, the better the accuracy and resolution.

Once all the samples representative of the pulse have been sampled, the sum of these samples is stored in a first adder 416;
This total sample value, ie, the integrated sum Σ of all sample values representative of the pulse, is stored in the second register 419. The second adder 41g and the second register 419 receiving its output contain the sum of intermediate values of the pulse sample values, that is, the value of double integration of the sample sequence. Once all samples of the pulse have been added, the addition and counting operations stop. The following three values are then stored. First, the count value N of the counter 422 is stored in the N register 426. The value of the second register 419 is then stored in the second register 428.

Finally, the value of the first adder 416 is added to the adder register 430.
be remembered. By substituting these three values into the following equation, the geometric mean η of each scanning pulse is obtained.

The classical centroid equation can be used in this embodiment, as well as in other embodiments where accurate pulse position determination is required. If the function indicating the magnitude of the pulse is f (t), the continuous form of the center of gravity equation is expressed by the following equation.

In a discrete system, the above equation becomes:

It has been confirmed that this centroid method is significantly advantageous over the previously described pulse integration method for pulses with a relatively small number of samples. The only problem with using this centroid method is that it has the same computational speed and economy as the pulse integral method.
The problem is that it is difficult to calculate the term ifl.

This problem can be solved by transforming the center-of-gravity equation into a form that does not require multiplication, thereby significantly reducing both calculation speed and cost. This modification method will be explained with reference to the data array shown in FIG. This data array is the following 3
It has two characteristics.

(1) Sum of all array elements: (n+i) ifl 00 (2) Sum of all array elements below the diagonal line: Σif1 0O (3) Sum of all array elements above the diagonal line: (n+x) Σf1 - Σif1 + Σ Σfj1 =0
14 1 Oj=0 Also, Σif1- (n+t)Σ fl-Σ Σ fj1=
0 1 Earth 0 1 Complement j-0 Therefore, the center of gravity equation can be transformed as shown below.

Here, the calculation for ifl 1o is obtained by adding the sample values of the pulses. Each intermediate value on the side of the array element above the diagonal line in FIG. 13 is obtained by the first adder 417 and sequentially stored in the register 430. Therefore, by simply summing these intermediate values, can be found. This value is obtained by the second adder 419 and stored in the second register 428. In this way, relatively inexpensive 2
Using only one adder, online calculations can be performed from one sample value to the next. To find the center of gravity of the pulse, after sampling is completed, division by is performed.

Once the geometric mean or centroid of each reference pulse has been determined with the required degree of resolution using one or more of the methods described above, the computer portion of the apparatus is compared with known parameters such as the distance between the optics and the jig to calculate the spatial coordinates of the workpiece.

When determining parameters necessary for calibration, one jig 9000 having a precise shape as shown in FIGS. 4 to 11 is used. This jig is used as a reference standard for performing calculations rather than as a stored spatial coordinate as used in the prior art. As shown in FIGS. 8 to 11, a plurality of pulse center points 901 scanned along the plane of the surface of the jig are selected, and a plurality of straight lines 902 connecting these points are formed.

The intersection points 904 of these straight lines thus formed determine the vertices of a geometric figure such as triangle 905 shown in FIGS. 9 and 10. The specific dimensions of the figure thus formed, ie the base and height of the triangle, are stored in the computer. The spatial coordinates of vertex 904 cannot be known by a computer. Further, these vertices do not need to be on the surface of the jig as shown in FIGS. 8 to 11, and it is preferable that the vertices are not on the surface of the jig.

The jig 900 shown in FIGS. 8 and 9 is a preferred form of jig because it is low cost, easy to mold, and does not collect foreign matter. The slots 908 formed in the jig 900 of FIGS.
It serves to dislodge other debris from within the jig, leaving the surface being scanned cleaner than other types of jigs. Furthermore, the jig 900 shown in FIGS. 8 and 9,
It can be easily and inexpensively formed on the surface of the support of the workpiece. Forming a jig on the workpiece support surface in this way also means that the jig does not support the workpiece. It also guarantees the relative position to the body with high precision, making calculations easier and improving accuracy.

Based on the information obtained by scanning the jig, the geometric figure unique to the jig is determined by the computer.
It is compared with each dimension of the figure stored in advance in the computer. One of the vertices of this geometric figure has coordinates 0,
It is defined as the origin of the three-dimensional coordinate system as the 0,0 point. The vertex selected as the origin of the coordinate system has its relative position with respect to the conveyor 16 determined, and this relative position is determined by the computer 2.
4, and in that coordinate system the position of each point on the surface of the workpiece is precisely determined. Of course, if the jig were formed on the workpiece support surface, ie the surface of the conveyor 16, the relative position of this jig would be fixed and preprogrammed into the computer. One or more other vertices of the geometry are determined in this coordinate system. Each point determined in this coordinate system, including the origin, will be used to calculate the calibration parameters set in the formula described later.

This unique method significantly improves the accuracy of the device according to the invention compared to conventional devices which are inevitably limited by the resolution of storing the spatial coordinates of the jig. According to the invention, calibration can be performed no matter where the optical scanner is pointed, as long as the jig is within the field of view of the camera. It is preferred, but not necessary, for the longitudinal axis of the jig to be oriented toward the longitudinal axis of the conveyor 16 to simplify calculation of the spatial coordinates of the workpiece.

Calibration means that calculates four parameters is used to calculate the YZ coordinates. The following equation is used to calculate the Y and two-space coordinates of each point on the workpiece.

Y"Kt CP-P' ) (Zo-2) -K2 Also, the calculation of the X space coordinate is performed using the following equation.

X giant Z tan β where η0. η. , Kll, 2 and Pl are the parameters calculated by calibration, η and P
is the parameter sent by the digitizer to the computer, zo is a constant equal to the distance between the camera and the calipresicon jig, and β is the angle of the laser with respect to the camera. A precisely shaped calibration jig and a video digitizer are used to sample data for each point on the surface of the jig, and each point in a three-dimensional coordinate system is determined as described above.

The four parameters (η0.η
~, Kl and 2) can be determined with high accuracy.

When using the jig shown in FIGS. 8-11, the computer is directed to the position of the Z, . This allows the system to select sample points on the flat surface of the jig. Two or more sample points 901 are selected on the flat surface of the jig (i.e., on the jig surface excluding intersections between the jig surfaces), and a line 902 connecting each of these points is formed by the computer. be done. And the intersection point 9 of a plurality of lines 902 formed from these selected points
04 is determined. Subsequently, each of these intersection points 904 is compared with a pre-stored dimension for a particular geometric figure, which is made up of visual points as each vertex of the figure. A coordinate system is taken with one of these vertices as the origin, and the remaining vertices are precisely defined under this coordinate system. These known points, including the origin, in this coordinate system are
It is used in the equation that transposes the equation regarding Y and 2 mentioned above, and solves these equations simultaneously (by η.
η one.

Ko and 2 are calculated. In this way, the device according to the invention allows calibration to be performed without storing the spatial coordinates of each point on the jig surface.

The four parameters can be determined periodically or continuously as described above.

This concludes the description of a novel and unique method and apparatus that were previously unknown and capable of determining the position of a workpiece at high speed and with high precision. However, it will be apparent to those skilled in the art that the present invention may be practiced in embodiments other than the specific embodiments described above without departing from the spirit and essential characteristics of the invention. The specific embodiments of high speed scanning devices described above are to be considered in all respects as illustrative only and not as limiting. Therefore, the gist of the present invention is not limited in any way by the embodiments described above, but is as described in the claims.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of a typical apparatus for determining the spatial coordinates of a workpiece according to the present invention, and FIG. 2 is a block diagram of an embodiment of a typical apparatus for determining the spatial coordinates of a workpiece according to the present invention using a camera with a pair of lenses. FIG. 3 is a block diagram of a typical video digitizer according to the present invention, and FIG. 4 is a typical calibration jig that can be used with the present invention. , FIG. 5 is a perspective view of the jig shown in FIG. 4, FIG. 6 is a side view of another embodiment of the calibration jig that can be used in the present invention, and FIG. 7 is a perspective view of the jig shown in FIG. FIG. 8 is a perspective view of an example of a preferred calibration jig formed on the support surface of a workpiece; FIG. 9 is a cut line 9-9 of the preferred calibration jig shown in FIG. 10 is a side view of a jig improved from the jig shown in FIG. 6, FIG. 11 is a perspective view of the jig shown in FIG. FIG. 13, which is a block diagram of another preferred example of a video digitizer, is a graph showing an arrangement of representative value data of digitized pulses. DESCRIPTION OF SYMBOLS 10...Laser, 12...Optical system, 14...Workpiece-16...Conveyor, 18...Photoelectric converter 7,
20...Optical system, 900...Jig. Applicant's agent Inomata Amendment of procedure (method) February/2nd, Showa Kaku, Director General of the Patent Office Kazuo Wakasugi 1, Indication of the case Showa Intercropping Patent Application No. 148882 2, Name of the invention Processed product Spatial coordinate determination method and device 3, relationship with the case of the person making the amendment Patent applicant Steven, James, White November 8, 1981 Drawing 8, Contents of the amendment Engraving of the drawing (no change in content)

Claims (1)

[Claims] 1. A method for determining the spatial coordinates of a workpiece, which comprises the following steps. (a) directing a planar beam of light from a light source onto the workpiece and the jig forming the unique geometry; (bl Memorizing the dimensions of the unique geometric figure of the jig. (0) At a position away from the workpiece, the jig, and the light source, find the intersection line of the planar light beam with the workpiece and the jig. A step of electrically inspecting and determining in advance the distance between this inspection position and the jig and the relative angle with the planar light beam. (d) A step in which the electrical images formed by the electrical inspection are (e) scanning along horizontal scan lines such that each scan line has a reference pulse corresponding to a point where the planar beam illuminates and intersects the workpiece or jig; (e) a selected scan; determining the center of gravity of the reference pulse on the line; (f) calculating a calibration parameter using the unique geometry of the jig; Calculating the spatial coordinates of the workpiece using the distance to the jig. 2. Before the step of electrically inspecting the intersection line between the planar light beam and the workpiece, the intersection between the planar light beam and the jig is calculated. A method for determining the spatial coordinates of a workpiece according to claim 1, comprising the steps of electrically inspecting the line, scanning the formed image, and calculating calibration parameters. 3. Calibration balance. 4. The method for determining spatial coordinates of a workpiece according to claim 1, wherein the step of calculating the coordinates of a plurality of workpieces is performed intermittently between calculations of spatial coordinates of a plurality of workpieces. A method for determining the spatial coordinates of a workpiece according to claim 1, comprising the additional step of using the spatial coordinates of the workpiece to indicate other devices acting on the workpiece. 5. A planar beam of light; scanning the electrical image formed by electrically inspecting the line of intersection between the workpiece and the jig includes sampling the signal of the electrical image and digitizing the sample value; A method for determining the spatial coordinates of a workpiece according to claim 1. 6. The step of determining the center of gravity of the reference pulse is performed by adding together the first sample value to the last sample value for the pulse. The width total values after each addition to obtain the first sum are added together to obtain a second sum, the second sum is divided by the first sum, and the last sample is obtained by dividing the second sum by the first sum. A method for determining the spatial coordinates of a workpiece according to claim 5, characterized in that the position of the specimen is returned by a number corresponding to the result of the division from . Method: (a) directing a planar beam of light from a light source onto the workpiece and a jig forming a unique geometry; (1)) Memorize the dimensions of the unique geometry of the jig (step. (e) At a position away from the workpiece, the jig, and the light source, The intersection line is electrically inspected, and the distance of this inspection position from the jig and the relative angle with the planar light beam are determined in advance (
step. (1) scanning the electrical image formed by the electrical inspection along a plurality of horizontal scanning lines, and generating reference pulses corresponding to points where the planar light beam illuminates and intersects the workpiece or jig; , each scan line has (+31) determining the geometric mean of the reference pulses on the selected scan line; (f) calculating a calibration parameter using the geometry specific to the jig; Step. (ω Calculating the spatial coordinates of the workpiece using the geometric mean of the reference pulse, the calibration parameter, and the distance between the inspection position and the jig. 8. Before the step of electrically inspecting the intersection line, a step of electrically inspecting the intersection line of the planar light beam and the jig, a step of scanning the formed image, and a step of calculating a calibration parameter are performed. A method for determining spatial coordinates of a workpiece according to claim 7. 9. Calculating calibration parameters comprises:
9. The method for determining the spatial coordinates of a workpiece according to claim 8, wherein the method is performed intermittently between calculations of the spatial coordinates of a plurality of workpieces. 10. A method for determining the spatial coordinates of a workpiece as claimed in claim 7, comprising the additional step of using the calculated spatial coordinates of the workpiece to direct other devices acting on the workpiece. 11. The step of scanning an electrical image formed by electrically observing the intersection line of the planar light beam, the workpiece, and the jig samples the signal of the electrical image, and calculates the sample value. 8. A method for determining spatial coordinates of a workpiece according to claim 7, comprising the step of digitizing. Claim 11 The step of determining the geometric mean of the reference pulses of the ratio-selected scan line comprises the steps of:
The method for determining the spatial coordinates of the workpiece described in Section 1. (a) Calculating the sum of all samples to determine the effective area of the lower portion of each pulse. (b) Integrate twice the value of each sample value in order from the first sample, and continue to integrate the values until this integrated value becomes equal to or exceeds the total value of all the samples. The step in which the last sample subjected to integration is used as the reference sample. (0) In the step of obtaining the reference sample, the interpolation element is calculated by calculating the ratio of the excess amount when the integrated value exceeds the total value of all the samples and the value twice the reference sample value. Deciding stage. ((1) adding the interpolation element to the reference sample position to perform interpolation on the reference sample position to determine the geometric mean of the pulses; 13. determining the geometric mean of the reference pulses of the selected scan line; A method for determining spatial coordinates of a workpiece according to claim 11, wherein the step of determining comprises the following steps: (a) calculating the sum of all samples to determine the effective area of the lower portion of each pulse; Calculating step. (1)) Integrate twice the value of each sample value in reverse order starting from the last sample, and whether this integrated value becomes equal to the total value of all the samples or the total value of all the samples. The step of continuing the integration until the value exceeds the value, and using the sample for which the last integration was performed as the reference sample. (0) determining an interpolation element by calculating a ratio between an excess amount when the integrated value exceeds the total value of all the samples and a value twice the reference sample value; ((1) subtracting the interpolation element from the reference sample position and applying it to the reference sample position to determine the geometric mean of the pulses; 14. Determining the spatial coordinates of the workpiece, comprising the following steps: Method: (a) A step of irradiating a planar light beam from a light source onto a jig forming a unique geometric figure. (1)) A step of memorizing the dimensions of the geometric figure. (C) Determining caliplane 1 parameters according to the following steps. (1) electrically inspecting the intersection line between the planar light beam and the jig from a known distance and a known angle with respect to the planar light beam; (2) electrically observing the intersection line between the planar light beam and the jig; (3) determining the center of the reference pulse; (4) storing the geometric figure; (d) irradiating the workpiece with a planar beam; (e) measuring the planar beam from the same position as the jig was inspected; (f) scanning an electrical image formed by the electrical inspection to correspond to a point where the planar light beam illuminates the workpiece; (h) determining the center of the reference pulse for the workpiece, the calibration parameter, and the jig and inspection position; calculating the spatial coordinates of the workpiece using the known distances; 15. scanning the electrical image formed by the intersection line of the planar light beam with the workpiece and the jig, comprising a plurality of horizontal scans; 15. The method of claim 14, comprising the step of scanning along a line to illuminate the workpiece or jig with the planar beam of light, such that each scan line has a reference pulse corresponding to its intersection point. 16. A method of determining spatial coordinates of a workpiece. 16. A method of determining spatial coordinates of a workpiece according to claim 15, wherein scanning the formed electrical image includes the step of digitizing the image. 17. , a jig forming a unique geometrical figure; a device for irradiating a workpiece and said jig with a planar light beam; a photoelectric conversion device located at a distance of , and located at a known angle with respect to the planar light beam; digitizing the output signal from the photoelectric conversion device and converting the light beam, the jig, and the processing to the light beam; a video digitizer device for determining the positions of the points of intersection with an object; the spatial coordinates of the workpiece using the calibration parameters, the positions of the intersection points of the light beam and the workpiece, and the known distance between the photoelectric conversion device and the jig; A computer device having a device for calculating , and a spatial coordinate determination device for a workpiece that scans the workpiece at high speed. 18. The device for determining the spatial coordinates of a workpiece according to claim 17, wherein the video digitizer is comprised of a digital circuit. 19. Claim 17, comprising an optical system having binocular lenses that form a pair of images on an image orthicon tube
A device for determining spatial coordinates of a workpiece as described in Section 3. 20. Claim 17, wherein the workpiece is supported by a substantially flat surface, and the jig is formed within the flat surface.
A device for determining spatial coordinates of a workpiece as described in Section 3. 21. The device for determining spatial coordinates of a workpiece according to claim 17, wherein the photoelectric conversion device is positioned at a known angle with respect to a plane formed by the planar light beam. 22. The process according to claim 1, wherein the jig is formed completely below the flat surface and has an opening that passes through the support of the workpiece, so that the jig can self-clean itself. A device for determining the spatial coordinates of objects. 23. The device for determining the spatial coordinates of a workpiece according to claim 22, wherein the flat surface is the upper surface of a conveyor for conveying the workpiece. 24. A workpiece is irradiated with a planar light beam, and a line of intersection between the planar light beam and the workpiece is set at a certain distance from the workpiece and at a known angle with respect to the planar light beam. A method for determining the spatial coordinates of a workpiece, which is inspected by a photoelectric conversion device that generates an electric signal, and which includes the following steps. (viewing) a step of placing a jig having a precise shape forming a unique geometric figure within the irradiation range of the planar light beam and within the field of view of the photoelectric conversion device; (bl: storing the dimensions of the unique geometric figure formed by the jig; +01: processing the electrical signals generated by the photoelectric conversion device in order to determine a group of intersections between the planar light beam and the jig; a) projection of the planar beam onto a flat surface of the jig;
selecting a group of points corresponding to . (θ) connecting the selected group of points by a plurality of straight lines and forming each vertex of the unique geometric figure of the jig by the intersection of the plurality of straight lines; (b) Defining a coordinate system based on each intersection point of the plurality of straight lines. @) Defining each intersection of the plurality of straight lines using the coordinate system. lhl Calculating calibration parameters using each of the intersection points defined by the coordinate system as a reference point. (il) generating a reference point corresponding to the projection of the planar beam onto the workpiece; Cj) setting the calibration parameter to x. applying equations regarding Y, Z coordinates to calculate the position of the reference point in the coordinate system and determining the spatial coordinates of each point on the surface of the workpiece; 25. irradiating the workpiece with a planar light beam; , a photoelectric conversion device that generates an electrical signal in which the line of intersection of the planar light beam and the workpiece is actually placed at a certain distance from the workpiece and at a certain angle with respect to the planar light beam; A method for determining the spatial coordinates of a workpiece, which includes the following steps: (al) Irradiating the planar light beam onto a jig having a precise shape forming a unique geometrical figure having several vertices. (bl) Memorizing the dimensions of the unique geometrical figure formed by the jig. Step. (C1 Step of scanning the jig with the photoelectric conversion device. Step of determining a group of intersections between the planar light beam and the jig as reference points. (C1 Step of scanning the jig with the photoelectric conversion device. Step of confirming the vertices. tj:l Step of defining a coordinate system using each of the vertices as a base. Q) Step of determining the position of each of the vertices in the coordinate system. (5) In the coordinate system, the position of each of the vertices A step of calculating a center axis calibration parameter based on the position. [1) A step of scanning the workpiece with the photoelectric conversion device. (3) calculating the position of the reference point on the workpiece in the coordinate system using the calibration parameters. The method for determining the spatial coordinates of a workpiece according to item 25. α) Selecting a group of intersection points between the planar light beam and the flat surface of the jig. 27. Each of C1 and (1) has the following steps: a step of sampling the signal level; a step of digitizing the sample values obtained in (o1 step); and each of steps 1(1) and (j) having the following steps: 1゜tp) Double the value of each sample value is accumulated in order from the first sample, and this accumulated value becomes equal to the total value of all the samples. Or, when the integrated value exceeds the total value of all the samples in the step of continuing the integration until it exceeds the total value of all the samples and using the last integrated sample as the reference sample. determining an interpolation element by calculating the ratio between the excess of 0 and the value twice the reference sample value; , determining the geometric mean of said pulses. 28. Each of the steps (C1 and (1)) has the following steps: (Jl) sampling the signal level generated by the photoelectric conversion device. (0) a step Q of digitizing the sample values obtained in step 1; and each of steps a) and (jl) comprises the following steps: Determination method: (pi: double the value of each sample is accumulated in reverse order starting from the last sample, and this accumulated value is either equal to the total value of all the samples or exceeds the total value of all the samples. The step of continuing the integration until the end of the integration and using the last integrated sample as the reference sample. (during stage q1), calculating the excess when the integrated value exceeds the total value of all the samples and the base S sample. Determining an interpolation factor by calculating the ratio of the value to twice the value of the pulse. Step of determining. 29. Sampling the pulse signal; digitizing the sample value as a result of this sampling to generate a digital quantity representing each of the sample values; and adding up each digital quantity in order to obtain a total sum. , and use this as the first sum, add the widths generated in the step of adding the digital quantities in order, obtain the sum, and use this as the second sum.
dividing the second sum by the first sum to determine an offset length, and determining the center of gravity of the pulse signal as a position back by the offset length from the last sample position of the pulse signal. determining the center of gravity of a pulse signal. 30. The method of determining the center of gravity of a pulse signal according to claim 29, wherein the first sum and the second sum are obtained using a digital adder. 31. electrically detecting a signal; integrating the signal by an analog integrator over one period of the signal including a pulse; and determining the first sum as a first sum; further integrating with an analog integrator and taking the result as a second sum, and dividing the second sum by the first sum, thereby determining the end position of one period of the signal and the pulse. A method for determining the center of gravity of a pulse signal, comprising the steps of creating a signal proportional to the temporal distance from the center of gravity.