JP2581155B2 - Absolute position detector - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an absolute position detecting device suitable for use in, for example, detecting the angle of an arm of an industrial robot or detecting the position of a machine tool when feeding a table.

[Summary of the Invention]

According to the present invention, a two-dimensional image sensor captures an image of a code plate formed with both a numerical pattern region and a stripe pattern region, and the arrangement relationship between the code plate and the image sensor has an inclination, so that a captured image is displayed. In an absolute position detection device, a step is generated at an edge of a stripe pattern, a coarse position is detected from an imaging output of a numerical pattern area, and a position with a high precision is detected from a step of the stripe pattern. By reading out only the line including the step portion necessary for detecting the absolute position from, the reading time of one frame is shortened, and the delay time of detecting the absolute position is shortened.

[Conventional technology]

Absolute position detectors detect angles of industrial robots,
It is often used for detecting a feed position of a table of a machine tool. The present applicant has previously filed Japanese Patent Application No. 62-19826.
As described in the specification, a numerical pattern is formed on a code plate serving as a scale plate, and a stripe-shaped pattern is formed. The code plate is read by a two-dimensional image pickup device constituted by a CCD, and imaged. An absolute position detecting device has been proposed in which the output signal of an element is processed to detect a coarse position from a numerical pattern and a fine position from a stripe pattern.

FIG. 4 shows a schematic configuration of the above-described absolute position detecting device, and includes a code plate 20 as a scale plate, and a code plate 20.
At least one of the image sensors 21 arranged to face the camera is movable. As shown in FIG. 5, the code plate 20 is provided with a numerical pattern area 22 indicating a gray code and a striped pattern area 23. Gray code is one of codes expressing binary numbers,
It is also called a binary code or a reflected binary code, and has a feature that only one bit always changes between two adjacent numbers. In the striped pattern region 23, a plurality of striped patterns P having a width w are formed in parallel with each other at equal intervals. The arrangement direction Z of these stripe patterns P is orthogonal to the extension direction of the pattern P,
The arrangement pitch is Pz. The gray code in the numerical pattern area 22 is assigned to each of the patterns P, and by reading this gray code, the absolute position in the Z direction can be known.

Image sensor 21 arranged opposite to code plate 20 described above
Is a two-dimensional image sensor using a CCD. The relative movement direction between the code plate 20 and the image sensor 21 is parallel to the arrangement direction Z. The axial direction of the imaging element 21 is provided so as to have a slight inclination angle θ with respect to the arrangement direction Z. In FIG. 5, the angle between the vertical axis V of the horizontal and vertical coordinate axes H and V of the image sensor 21 and the arrangement direction Z is represented by θ.

The imaging element 21 simultaneously captures a part of each of the pattern areas 22 and 23 of the code plate 20 and sends an imaging signal to the detection circuit 24 in FIG. , The value of the gray code is read from
The coordinates of the predetermined point are read. These read data are supplied to the coordinate conversion circuit 25. The coordinate conversion circuit 25
The absolute position information in the Z-axis direction is calculated based on the value of the gray code and the coordinates of a predetermined point of the stripe pattern P. That is, the gray code provides relatively coarse position information in units of the array pitch Pz of the pattern P, and the coordinate conversion based on the coordinates of a predetermined point of the pattern P provides higher position information. .

In addition, in FIG.
Illustrates an example in which light from a light source 21 is used and a so-called transmissive code plate 20 is used, but a reflective code plate in which a light source is arranged on the same side as the imaging element 21 may be used. .

FIG. 7 shows a part of an image taken by the image pickup device 21 in an enlarged manner. The pattern P in the stripe-shaped pattern region 23 is a binarized imaging pattern for each pixel (cell) S of the imaging element 21. In this case, V of the image sensor 21
Since the axial direction and the arrangement direction Z of the stripe pattern P have an angle θ, the edge E of the pattern P has a stepped shape as shown in FIG. This step boundary E is
It occurs at the upper edge Ea and the lower edge Eb of the pattern P. The pixels Sa1, Sa2, etc. of the step portion of the upper edge Ea can be detected as pixels that change from white to black when scanning in the H direction, and the pixels Sb1, Sb2, etc. of the step portion of the lower edge Eb in the H direction. It can be detected as a pixel that changes from black to white when scanning.

Here, when the coordinates of the image sensor 21 in the directions of the H axis and the V axis are expressed in units of the number of pixels S, the coordinates of the pixel Sa1 at a predetermined step portion of the upper edge Ea become (Ha1, Va1). The coordinates of the pixel Sb1 at the step portion of the lower edge Eb corresponding to the pixel Sa1 are (Hb
1, Vb1). The coordinates (H1, V1) of the center position of the line connecting these pixels Sa1 and Sb1 are as follows: H1 = (Ha1 + Hb1) / 2 V1 = (Va1 + Vb1) / 2.

The point of the coordinates (H1, V1) is on the center line c of the stripe pattern P as shown in FIG. 7, and the absolute position Zc of the pattern center line c in the Z direction can be read from the gray code. Can be. Moreover, each coordinate value, because they represented the number of pixels S units, when each pitch on the H axis and the V axis direction of the pixel S to respectively P _H and P _V, a position read reference point of the image sensor 21 Coordinates (H for the origin (0,0)
The actual distance of (1, V1) is H1P _{H in the} H-axis direction and V1P _{V in the} V-axis direction. Therefore, the absolute position Z ₀ of the origin (0, 0) in the Z-axis direction is the absolute position Zc of the pattern P, and the coordinate values H1, V
1. From the pitches P _H and P _{V of} the pixel S and the oblique viewing angle θ, it can be obtained as Z ₀ = Zc− ( _H 1 P _H sin θ + _V 1 P _V cos θ). Thereby, in principle, the absolute position can be read with the accuracy of the minimum unit P _H sin θ.

Note that the coordinates of each pixel S are all represented by integer values,
The coordinates (H1, V1) of the center position and the like are integer multiples of 0.5.

[Problems to be solved by the invention]

In the above-described absolute position detection device, since the image sensor 21 reads all pixels, the reading time of one frame becomes longer,
There was a problem that the output delay time of the current position information became long.
If the output delay time is long, the use range is limited. The output delay time is represented by the following equation.

(Output delay time) = 1/2 (Exposure time) + (Pixel signal transfer processing communication time) When the image sensor 21 is of the interline system, the charge readout order of each pixel is as follows.

The charge of the light sensor diode is transferred to the vertical shift register.

Execute the shift operation of the vertical shift register. At the same time, the first horizontal line charge is transferred to the horizontal shift register.

Execute the shift operation of the horizontal shift register. At the same time, the first charge is read.

The above operations are performed in the order shown in FIG. In FIG. 8, Nh indicates the number of horizontal pixels, and Nv indicates the number of vertical pixels. Assuming that the shift time required for the operation is T1, the shift time required for the operation is T2, and the shift time required for the operation is T3, T1 is sufficiently shorter than T2 and T3, and can be ignored. The time T required to read out the charges of all the pixels of the element 21 is T ≒ (T2 + T3 × Nh) · Nv.

As an example, the horizontal shift operation frequency is 10 MHz (T3 = 0.1
μsec) and the vertical shift operation frequency is 1MHz (T2 = 1μse
In c), when 90,000 pixels of (300 × 300) are read, 1
The frame read time T is about 9.3 mse
becomes c.

(1 .mu.sec + 0.1 .mu.sec.times.300) = 9.3 msec By the way, focusing on the image of the image pickup device 21, a portion providing information necessary for outputting the absolute position information is a very small portion such as an edge of the stripe pattern P. That is, at the edge of the stripe pattern P, the coordinate value of a point at which the color changes from white to black or from black to white, and the value of pixel data (0 or 1) near the center of each track of the numerical pattern on the horizontal line having the change point ) Is required.

Accordingly, it is an object of the present invention to provide an absolute position detecting device in which reading time of one frame is shortened by reading only necessary pixels instead of reading out all pixels, and the responsiveness is improved.

[Means for solving the problem]

According to the present invention, at the exit of the transfer registers 2D and 2U connected to the horizontal shift register 3 of the two-dimensional image pickup device 21, a discriminating circuit 16 for discriminating the necessity or unnecessary of the signal charge is provided. When the signal charge is required, the vertical shift operation is performed after confirming that the horizontal shift register 3 is in the receiving state, and the top charge is transferred to the horizontal shift register 3 and the address of the horizontal line is held. According to the output of the discriminating circuit 16, when the signal charge is unnecessary, the vertical shift operation is executed to sweep out the head charge.

[Action]

When reading the stripe pattern area of the two-dimensional image sensor 21, information necessary for detecting the absolute position is included in the line of the step. Therefore, if only the line including the step portion is read, the reading time of one frame of the image sensor can be reduced as compared with reading all the pixels. Whether or not the stepped portion is included can be detected by the discrimination circuit 16 from the coincidence and non-coincidence of the binarized imaging signals of the pixels at both ends of the stripe pattern. In other words, if they match, the line does not include a stepped portion and does not need to be read. If they do not match, it is a line that needs to be read. The electric charge of the line that does not need to be read is not supplied to the horizontal shift register 3 and is discarded through the ground line. The charge on the line that needs to be read is
It is supplied to the horizontal shift register 3 and taken out to the output terminal by the horizontal shift operation.

〔Example〕

Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In this embodiment, the present invention is applied to the previously proposed absolute position detecting device.

In this absolute position detecting device, an image showing a part of the image is obtained in FIG. In FIG. 2, x represents a step portion formed at the edge of the pattern P in the stripe-shaped pattern region 23,
Indicates a reading point of the numerical pattern area 22. G1, G2, G3
.. G10 indicate gray code values obtained by collecting pixel data values (0 or 1) near the center of each track. In the example shown in FIG. 2, the change point coordinate values and the gray code values corresponding to the ten sets of step portions are obtained, and ten pieces of absolute position information are obtained by the above-described calculation. That is, to obtain the absolute position information, only the output signal of the horizontal line having the coordinate change point is required.

FIG. 1 shows a configuration for reading out only horizontal lines having necessary information. In FIG. 1, the signal charge from the photodiode 1 shaded in FIG.
・ Transferred to 2D ・ ・ 2E. Vertical shift registers 2U and 2D
Is for transferring the signal charges of the pixels 23D and 23U at the left and right ends of the pattern P of the stripe pattern area 23 in FIG. 2, and the vertical shift register 23E corresponds to the rightmost pixel of the numerical pattern area 22. ing.

The leading signal charges of the vertical shift registers other than the vertical shift registers 2D and 2U are transferred to the horizontal shift register 3 line by line or are swept out through the ground line 4. In FIG. 1, a circuit for switching between transfer to the horizontal shift register 3 and sweeping through the ground line 4 is omitted. The output charge of the horizontal shift register 3 is converted into a signal voltage by a charge-voltage conversion circuit 5 and supplied to a comparison circuit 6. The threshold voltage is supplied to the comparison circuit 6 from the terminal 7, and a binarized output signal is obtained at the output terminal 8 of the comparison circuit 6.

The first charge of the vertical shift register 2D is a charge-voltage converter
The head charge of the vertical shift register 2U is supplied to the comparison circuit 10D via the charge-voltage conversion circuit 9U via the charge-voltage conversion circuit 9U.
Supplied to 10U. These comparison circuits 10D and 10U include:
A threshold voltage is supplied from the terminal 7. Output signals (0 or 1) of the comparison circuits 10D and 10U are supplied to the exclusive OR gate 11. The output signal of the exclusive OR gate 11 is supplied to the AND gate 12, and the AND gate 12
The 12 output signals are supplied to the OR gate 14. AND gate 12
, A read inhibit signal is supplied from the terminal 13, and a forced read signal is supplied from the terminal 15 to the OR gate 14.

The output signal of the OR gate 14 is supplied to the determination circuit 16, and when the output signal of the OR gate 14 is 1, the vertical shift register 2D
When the output signal of the OR gate 14 is 0, the signal charge of the line including the top charges of the vertical shift registers 2D and 2U is transferred to the horizontal shift register 3. Are swept out through the ground line 4 and discarded.

That is, the presence or absence of a coordinate change point is determined by the stripe pattern area.
It can be determined from the binarized values of the pixels 23D and 23U at the left and right ends of 23. If the values of the pixels in these two columns are different,
If there is a coordinate change point and the values of both pixels are the same, it can be determined that there is no coordinate change point. If the determination circuit 16 determines that there is no coordinate change point, the signal charges on that line are discarded. On the other hand, if the determination circuit 16 determines that there is a coordinate change point, the operation of the horizontal shift register 3 is stopped (an acceptable state), and the vertical shift operation is executed. Transfer to At the same time, the vertical address of the line determined to have a coordinate change point is stored in a vertical address register (not shown).
Is forwarded to

When the read inhibit signal supplied from the terminal 13 to the AND gate 12 is set to 0, the read is forcibly inhibited regardless of the presence or absence of the coordinate change point. This read prohibition operation is used to limit the number of lines to be read and suppress the read time of one frame within a predetermined time. If the forced read signal supplied from the terminal 15 to the OR gate 14 is 1, all pixels are forcibly read regardless of the presence or absence of a coordinate change point. This forced read operation is used for quality inspection of the image sensor 21 and the like.

When the image sensor 21 is of the interline type, the order of reading out the charges of each pixel is as follows.

The charge of the light sensor diode is transferred to the vertical shift register.

Execute the shift operation of the vertical shift register. At the same time, the first horizontal line charge is transferred to the horizontal shift register.

Execute the shift operation of the horizontal shift register. At the same time, the first charge is read.

In one embodiment of the present invention, the above operations are performed in the order shown in FIG. The first vertical shift operation is an operation performed until a coordinate change point is first detected. When the coordinate change point reaches a certain line, a transfer operation through the horizontal shift register 3 is performed. In parallel with this, a vertical shift operation is performed, and a line having the next coordinate change point is detected.

The shift time required for the operation is T2 (for example, 1 μs), and the shift time required for the operation is T3 (for example, 0.1 μs).
s) and the number of horizontal and vertical pixels is 100, C
The time T required to read out the charge of the pixel having the necessary information of the CD image sensor 21 is T 0-9 + (1 + 0.1 × 100) × 10 = 110-119 μs. Therefore, the readout time for one frame is shorter than when all the pixels of the image sensor 21 are read out.

The present invention can be applied to a CCD image sensor other than the interline type.

〔The invention's effect〕

According to the present invention, since only lines including pixels necessary for obtaining the absolute position information are continuously read, the readout time of one frame is shortened as compared with reading out all the pixels of the image sensor, and the output delay is reduced. Time can be greatly reduced.
Further, if the read time is set to be the same as the time for reading all the pixels, there is an advantage that the clock frequency of the shift operation is reduced and the power consumption is reduced.

[Brief description of the drawings]

FIG. 1 is a connection diagram of one embodiment of the present invention, FIG. 2 is an enlarged view showing a part of an image of an image pickup device, FIG. 3 is a schematic diagram used for explaining the operation of one embodiment of the present invention. FIG. 4 is a block diagram showing a schematic configuration of an absolute position detecting device to which the present invention can be applied, FIG. 5 is a plan view showing an example of a code plate pattern,
FIG. 6 is an enlarged plan view showing a stripe pattern P imaged by the image sensor, FIG. 7 is a schematic diagram for explaining a coordinate conversion operation, and FIG. 8 is a diagram for explaining a read operation of the image sensor. FIG. Description of main symbols in the drawings 2D, 2U, 2E: vertical shift register, 3: horizontal shift register, 4: ground line, 11: exclusive OR gate, 16: judgment circuit, 21: image sensor, 22: numerical pattern area, 23: Stripe pattern area.

Claims (2)

(57) [Claims] An output circuit of a transfer register connected to a horizontal shift register of a two-dimensional imaging device is provided with a determination circuit for determining whether signal charge is required or not. At the time
After confirming that the horizontal shift register is in the accepting state, a vertical shift operation is performed to transfer the top charge to the horizontal shift register, hold the address of the horizontal line, and output the signal from the determination circuit. An absolute position detecting device, wherein the vertical shift operation is performed when no charge is required, and the head charge is swept out. 2. A patent wherein a control circuit which forcibly requires and eliminates signal charges from the outside is provided on the output side of the discriminating circuit, and a forced reading function or a reading prohibiting function is added. The absolute position detecting device according to claim 1.