JPH01298479A - Inclination angle detector - Google Patents

Abstract

PURPOSE:To detect the inclination angle of a high speed, stable and arbitrary side by setting a window for the arbitrary side of an object picture, equally dividing an area in the window and calculating the inclination angle of the side from the area of the object part in respective dividing areas. CONSTITUTION:A window 4 is set at an arbitrary side 3 for an object picture 2 obtained by image-picking up an object, the area in the window is equally divided, the area of the object part in respective dividing area is measured, the inclination angle of the side 3 is calculated from the measured result and the posture, etc., of the object are detected. In this case, the number of picture elements to constitute the object part of the picture is counted from the video signal, the area is measured, the vertical flyback period of the video signal is used, the calculation operation of the inclination angle is executed, and the inclination angle of the side perceived by the video rate can be detected. Since it is sufficient when a single window is set, the operability is excellent, and since the integrating processing is executed, the measurement error due to the noise and the distortion of the picture is smaller and the stable measurement can be performed.

Description

[Detailed Description of the Invention] <Industrial Application Field> The present invention relates to a tilt angle detection method that detects the tilt and posture of an object by measuring the tilt angle of any side of an object image having a polygonal shape, for example. Regarding equipment.

<Prior art> Conventionally, this type of inclination angle detection device performs predetermined image processing on an object image obtained by a camera to determine the outline of the object part, or detects the outline of the object part or the end points of the side of interest. After detecting two arbitrary points, the inclination angle of the side of interest is detected from these contour lines and detection points.

<Problems to be Solved by the Invention> However, the method using contour lines complicates image processing and takes a long time, and the method using detection points makes it difficult to detect end points or arbitrary two points. Furthermore, there are problems such as increased detection errors due to the effects of noise and image distortion.

This invention was made by focusing on the above problem, and after setting a window for any side of an object image, it performs integral processing on the image within that window in real time. It is an object of the present invention to provide a novel tilt angle detection device that can stably detect the tilt angle of any side.

Means for Solving the Problems In order to achieve the above object, the inclination angle detection device of the present invention includes means for capturing an image of an object to generate an object image, and setting a window for any side of the object image. a means for equally dividing the area within the set window; a means for measuring the area of the object part included in each divided area; and a means for setting the window based on the area of the object part in each divided area. It was decided to provide a means for calculating the inclination angle of each side.

Effect> To detect the posture of an object, window settings are performed by focusing on an arbitrary side of the object image obtained by capturing the object. After that, the area within the window is divided equally, the area of the object part in each divided area is measured, and the inclination angle of the side of interest is calculated from the measurement results to detect the posture of the object. become. In this case, if you count the number of pixels that make up the object part of the image from the video signal, measure the area, and calculate the tilt angle using the vertical blanking period of the video signal, you can The angle of inclination of the side can be detected. In addition, it is easy to operate because you only need to set up a single window, and because it uses integral processing,
Measurement errors due to noise and image distortion are reduced, allowing stable measurements.

<Embodiment> FIG. 1 shows an example of the circuit configuration of a tilt angle detection device according to an embodiment of the present invention, and FIGS. 2 and 3 show the principle of this device.

FIGS. 2 and 3 show an object image 1 expressed in an XY coordinate system, and this image includes a polygonal object portion 2. FIG. Specifically, this object image 1 is a binary image obtained by binarizing a video signal obtained by imaging an object, and the object portion 2 is, for example, a black pixel, and the background portion 2' is a binary image. Consists of white pixels.

In Fig. 2, we focus on a side 3 of the object part 2 that has an almost horizontal inclination, and two points P (
A rectangular window 4 is set by specifying X,,Y,) and Q(χ2°Yz).

If the area within this window 4 is divided equally on the left and right by the central vertical line 6, the length of the window 4 in the X direction is
−X, so each rectangular divided area a,
The length of b in the X direction! ! ,,f becomes (xz-x,)/2.

Now, each divided area a on the left and right across the vertical line 6.

The area (specifically, the number of black pixels) of the object portion 2 included in b is set as S, ,S, and each divided area a is divided into two areas.

If the intersections of center lines 7 and 8 of b and side 3 are A and B,
Displacements ΔX and ΔY in the X and Y directions between intersections A and B
is given by the following equation.

X, -X.

ΔX=□ ...■ Also, the angle formed by side AB with respect to the horizontal line 5 passing through point A is θ
′, then , and the tilt angle θ and angle θ′ are θ=π−θ′
Since the following relationship holds true, the inclination angle θ is expressed as follows from the above equations 0 to 0.

Next, in FIG. 3, we focus on the side 9 of the object part 2 that has an almost vertical inclination, and with respect to this side 9, two points R (X3, Y3) and T (x,,y,) By specifying , a rectangular window 10 is set.

The area within the window 10 is equally divided vertically by the central horizontal line 12, and the area (specifically, the number of black pixels) of the object portion 2 included in each of the upper and lower divided areas c and d with this horizontal line 12 in between is calculated. are Sc and Sd, and the intersections of the center lines 11.13 of each divided area c and d with the side 9 are C and D, then the displacements ΔX and ΔY in the X direction and the X direction between the intersections C and D are , as before, is given by the following equation 00, and the inclination angle θ is given by the following (2).

In this example, for each horizontal scanning line of the video signal, the area (
The number of black pixels) is determined, and the cumulative addition value of the area (number of black pixels) over the effective vertical scanning period is determined for each horizontal scanning line, and the formula 0 or 0 is calculated based on the cumulative addition value during the next vertical retrace period. The calculation of the formula is executed to calculate the inclination angle θ of the side of interest.

FIG. 4 shows a time chart of the video signal VD in this embodiment. In the figure, the upper half of the diagram shows the video signal VD during the vertical scanning period, and the lower half of the diagram shows the video signal VD during the horizontal scanning period.

In the figure, VD indicates a vertical synchronization signal, HD indicates a horizontal synchronization signal, and one vertical scanning period (16.7 milliseconds) is 20H (however, LH means one horizontal scanning period, which is 63.5 microseconds). , and the remaining valid vertical scanning period. This effective vertical scanning period includes 242 horizontal scanning lines, and each horizontal scanning line includes:
It contains 256 pixel data.

FIG. 1 shows a specific example of a tilt angle detection device based on the above principle.

In FIG. 1, a television camera 20 images an object 21 from above, for example, and generates a grayscale image. The video signal VD of this grayscale image is output to the synchronization separation circuit 22, where it is output as a horizontal synchronization signal 11D.

Vertical synchronization signal VD, clock signal CK, etc. are separated. The binarization circuit 23 sets a certain threshold level for the video signal νD to binarize the grayscale image into black and white to generate a binary image. The window control circuit 24 uses the window setting unit 40 to set two points P,
When Q, R, and T are specified and window 4.10 is set, the X direction area
, ~X2 and X3 ~ x4, the window signal WD is turned on to control the opening and closing of the gate circuits 25 and 26.

These gate circuits 25 and 26 have two
One gate circuit 25 gives binary data to the first pixel counter 27 when the window signal WD is on, and the other gate circuit 26 gives the binary data to the first pixel counter 27 when the window signal WD and the division signal DV (described later) are on. At this time, binary data is given to the second pixel counter 28.

The first pixel counter 27 receives the video signal VD.

For each horizontal scanning line of
The second pixel counter 28 is used to count the number of black pixels included in one divided area within the window 4, and each pixel counter 27 , 28 count value N4. N, 2 is I10
CP in the microcomputer via port 29
Incorporated into U30. In this case, the CPU 30 will take in the count values corresponding to the Y-direction areas Y, -Y2, and Y3 to Y4 of the window 4.

The horizontal counter 31 is for giving the horizontal coordinate address (X coordinate in Figures 2 and 3) of the binary image, and the latch circuit 34 is for giving the horizontal coordinate position of the vertical line 6 that equally divides the window left and right (see Figure 2). Then (X+ +Xz) / 2)
The comparator 33 is for comparing the counted value of the horizontal counter 31 and the latch data of the latch circuit 34 to determine whether they match or do not match. The matching output and the horizontal synchronizing signal HD are manually input to output a divided signal DV that turns on corresponding to the divided area.

The vertical counter 36 is for giving the vertical coordinate address (Y coordinate in FIG. 2) of the binary image, and starts counting the horizontal synchronizing signal +1D after a time corresponding to 20H has elapsed since the fall of the vertical synchronizing signal VD. , and its count value Y,
is taken into the CPU 30 via the I10 boat 29.

When the CPU 30 is based on the principle shown in FIG.
Each time there is an interrupt TNT caused by the horizontal synchronizing signal +10 for 30, the control procedure shown in FIG. 5 is executed to detect the inclination angle θ of the side 3 of interest. Further, when based on the principle shown in FIG. 3, the CPU 30 similarly executes the control procedure shown in FIG. 6 every time there is an interrupt INT caused by the horizontal synchronization signal HD, and detects the inclination angle θ of the side 9 of interest. do.

In FIG. 1, FROM 37 stores a series of programs related to detecting the tilt angle θ, and RAM 38 stores various data and is also used as a work area. Further, the communication control unit 39 transmits the calculation results of the CPU 30 to a higher-level controller.

Next, the operation of the circuit shown in FIG. 1 will be explained in detail based on FIG. 5.

First, when the object 21 is imaged by the television camera 20, the synchronization separation circuit 22 generates a vertical synchronization signal VD.

The horizontal synchronization signal 11D and the clock signal CK are separated, and the video signal VD is further binarized in the binarization circuit 23 to generate a binary image.

This binary image output is passed through gate circuits 25 and 26 to the first . The first pixel counter 27 counts the black pixels included in the window 4 for each horizontal scanning line when the gate circuit 25 is turned on by the window signal WD. Further, the second pixel counter 28 counts the number of black pixels included in the divided area when the gate circuit 26 is turned on by the window signal WD and the divided signal DV.

Then, when an interrupt INT is generated by the horizontal synchronization signal HD to the CPU 30 in each horizontal scanning period, first the CPU
30 is step 1 in FIG. 5 (indicated by rsTIJ in the figure)
In step 3, the CPU 30 reads the count value N4 of the first pixel counter 27, and in step 2 calculates the cumulative addition value N and stores it in the RAM 3B. Count value N
, 2 is read, and in the subsequent step 4, the cumulative addition value N2 is read.
is calculated and stored in the RAM 38.

Next, in step 5, the CPU 30 reads the count value Y of the vertical counter 36, and in the following step 6, the CPU 30 reads the count value Y,
It is determined whether or not has reached 242, that is, whether or not the processing of steps 1 to 5 for the final horizontal scanning line has been completed, and if the determination is "NO'", the next interrupt TNT is executed. The process waits, and the above-described processes for the following horizontal scanning lines are repeatedly executed.

When the processing for the final horizontal scanning line is thus completed, the step 6 becomes "YES" and the process proceeds to step 7, where the CPU 30 calculates the cumulative addition value N (included in the window 4) of the count value of the first pixel counter 27. By subtracting the cumulative addition value N2 of the count value of the second pixel counter 28 (corresponding to the area S of the object part 2 included in the divided area A) from the divided area a (corresponding to the area of the object part 2),
A value N1 corresponding to eight areas of the object portion 2 included in is calculated and stored in the RAM 38.

Next, in step 8, the CPU 30 controls each divided area a.
, b by using the area S, , S and the setting data from the window setting unit 40 to calculate the inclination angle θ, and finally the data N stored in the RAM 3B.

Clear N, , N2 (step 9).

Next, when the principle shown in FIG. 3 is used, the control procedure shown in the sixth time will be executed. In this case, when an interrupt [NT by the horizontal synchronizing signal HD occurs to the CPU 30 in each horizontal scanning period, the CPU 30 first calculates the count value N of the first pixel counter 27 in step 11 of FIG.

In the following step 12, the count value Y of the vertical counter 36 is read.

Next, in step 13, the CPU 30 starts the vertical counter 36.
The count value Y, is the boundary position (Y 3 +
It is determined whether Y 4 ) / 2 has been reached.
If the determination is "NO," the process proceeds to step 14, where the CPU 30 executes the cumulative calculation of the count value NJ, and stores the calculated value as the area equivalent value N of the divided area C in the RAM 3B.
stored in a predetermined area.

If the determination in step 13 is "YES", it is determined that the measurement of the area of divided region C has been completed, and step 1
5, the CPU 30 similarly performs the cumulative calculation of the count value NJ, and stores the calculated value in another area of the RAM 38 as the area equivalent value N2 of the divided region d.

Next, in step 16, the CPU 30 determines whether the count value Y of the vertical counter 36 has reached 242, that is, whether or not the processing of steps 11 to 15 for the final horizontal scanning line has been completed. If the judgment is “
If NO°, the process waits for the next interrupt INT, and the above-described process is repeatedly executed for the following horizontal scanning lines.

When the processing for the final horizontal scanning line is thus completed, step 16 becomes "YES" and step 17 is executed.
Then, the CPU 30 calculates two cumulative addition values N, (corresponding to the area Sc of the object portion 2 included in the divided region C) and NZ (the area S, of the object portion 2 included in the divided region d).
) and the setting data from the window setting unit 40 to calculate the angle θ around 1 by executing the calculation of the above equation 0, and finally clear the data N, , N2 stored in the RAM 38. (Step 18).

In the above embodiment, the first pixel counter 27 calculates the area of the object part 2 included in the window 4, and the second pixel counter 28 calculates the area S of the object part 2 included in the divided area. , the area S of the object part 2 included in the divided area a is calculated from the difference between the two, but the area of the object part 2 included in each divided area a and b is calculated directly using two pixel counters. It is also possible to

FIG. 7 shows a specific example of this, in which a flip-flop 41, two gate circuits 42, a 43° pixel counter 44
．． The area S of the object portion 2 included in the divided area is determined using the circuit group 51 consisting of the latch circuit 45, and the area S of the object portion 2 included in the divided area is calculated, and the gate circuits 47, 48, 446. Pixel counter 49. The area S of the object portion 2 included in the divided area a is determined using a circuit group 51 consisting of a latch circuit 50.

<Effects of the Invention> As described above, the present invention sets a window for any side of an object image, divides the area within the window equally, measures the area of the object part in each divided area, Since the configuration is configured to calculate the inclination angle of the side of interest from the measurement results, the area is measured by counting the number of image components of the object part from the video signal, and the inclination angle is calculated using the vertical retrace period of the video signal. It is possible to calculate the angle of inclination of any side using video rate. In addition, since it is sufficient to set a single window, it has excellent operability, and since it uses integral processing, measurement errors due to noise and image distortion are small, and stable measurement is possible, achieving the two objectives of the invention. It has a remarkable effect.

[Brief explanation of the drawing]

FIG. 1 is a circuit block diagram of a tilt angle detection device according to an embodiment of the present invention, FIGS. 2 and 3 are explanatory diagrams showing the principle of the invention, FIG. 4 is a time chart of a video signal, and FIG. 6 and 6 are flowcharts showing the procedure of CPU interrupt control, and FIG. 7 is a circuit block diagram showing another embodiment. 4.10... Window 20... Television camera 24... Window control circuit 25, 26... Gate circuit 27, 28... Pixel counter 30... CPU 31. ... Horizontal counter 33 ... Comparator 34 ... Launch circuit 36011. Vertical counter 40...Window setting section

Claims (1)

[Claims] An inclination angle detection device for detecting the inclination angle of an arbitrary side of an object image, comprising means for capturing an image of an object to generate an object image; A means for setting a window, a means for equally dividing an area within the set window, a means for measuring the area of an object part included in each divided area, and a means for measuring the area of an object part within each area. A tilt angle detection device comprising means for calculating a tilt angle of a side on which a window is set.