JPH0528268A - Image processing device - Google Patents

Abstract

(57) [Summary] [Object] To provide an image processing apparatus capable of detecting the position and inclination of a large rectangular work that does not fit within the field of view of a CCD camera or a highly accurate rectangular work. [Structure] The first and second two CCD cameras and a position / tilt detecting means for detecting the position and inclination of a square work are provided, and two CCD cameras respectively capture images of two vertices of the square work. Based on the image data of the two vertices, the coordinates of the two vertices are obtained by the binarization process, the run length address creation process, the circumscribed quadrangle extraction process, and the vertex calculation process, and the quadrangle work of the quadrangle work is calculated from these two vertex coordinates. Detect position and tilt.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus, and more particularly to an image processing for detecting the position and inclination of a quadrangle work by digitally processing image data of the quadrangle work obtained by an optical method. It relates to the device.

[0002]

2. Description of the Related Art In a conventional image processing apparatus, one CCD
The whole image of the quadrangle work is taken with a camera, the obtained image data is taken in and stored in the image memory, and it is binarized by software or hardware.
The position and the inclination of the quadrangle work are calculated by calculating the barycentric coordinates and the secondary moments in the X and Y coordinate axis directions from each of the binarized pixels.

This conventional technique will be described with reference to the drawings. First, the position detection will be described. In the conventional position detection, an image is read by one CCD camera or the like, and after binarization processing, calculation is performed from image data to obtain barycentric coordinates. A conventional technique will be described with reference to FIG. Figure 9
FIG. 7 is an explanatory diagram showing a binary image by a conventional image processing device.

In the figure, the shaded area indicates existing pixels, and the entire shaded area, which is a group of the pixels, is an image. Then, in order to detect the position from this image, the barycentric coordinates are generally obtained. Figure 1 shows how to find this center of gravity.
0 and the flowchart of FIG. 11 will be described. Figure 1
0 is a flowchart showing a process of obtaining the X-coordinate of the center of gravity by the conventional image processing apparatus, and FIG. 11 is a flowchart showing a process of obtaining the Y-coordinate of the center of gravity by the conventional image processing apparatus. By performing the processes shown in FIGS. 10 and 11 on all the pixels of the pixel block on the coordinate screen as shown in FIG. 9, the barycentric coordinates of the image as the pixel block can be obtained.

First, a method of calculating the X coordinate will be described. In FIG. 10, in step S1001, the line number n in the X direction is initialized. Then, step S10
At 02, the start and end coordinates (X
s, Xe), and n in the X direction is calculated in step S1003.
The number of pixels of the line (pixel) is calculated, and step S10
In 04, the number of pixels of the n line (pixel) is added to the number of pixels before the n line to obtain the total number of pixels (area). In step S1005, grv is calculated by the following equation (1). grv = grv + {pixel * (Xs + Xe)} / 2 (1) This grv can be understood as an area moment (first moment) about the Y axis by the pixel.

Thereafter, in step S1006, it is determined whether or not the calculation of the grv of the last line is completed. If the calculation of the grv of the last line is not completed yet, the calculation of the line in the X direction is completed in step S1007. The line number n is incremented, the process returns to step S1002, and the above process is repeated for the next unprocessed line. Then, after the calculation of the grv of the last line is completed, step S1
Proceeding to 008, the X coordinate of the center of gravity is obtained by the following equation (2). gravX = grv / area (2) Expression

Here, the X coordinate of the center of gravity of the binary image of FIG. 9 is calculated by the above method. First, in the processing of the first line, since there are no pixels and the number of pixels is 0, grv is also 0. In the processing of the second line, Xs = 1, Xe = 2, pixel
Since l = 2 and area = 2, grv = 0 + {2 *
(1 + 2)} / 2 = 3. In the processing of the third line, Xs = 1, Xe = 2, pixel
Since l = 2 and area = 4, grv = 3 + {2 *
(1 + 2)} / 2 = 6. In the processing of the fourth line, Xs = 1, Xe = 4, pixel
Since l = 4 and area = 8, grv = 6 + {4 *
(1 + 4)} / 2 = 16. In the processing of the fifth line, Xs = 1, Xe = 4, pixel
Since l = 4 and area = 12, grv = 16 +
{4 * (1 + 4)} / 2 = 26. This completes the processing for all the lines, and the processing for obtaining the X coordinate gravX of the center of gravity is performed using this result. That is, gravX = 26/12 = 2.166 ..., and the X coordinate of the center of gravity of the binary image in FIG.
It becomes 2.

Next, a method of calculating the Y coordinate will be described. In FIG. 11, in step S1101, the line number n in the X direction is initialized. Then, step S11
02, the start and end coordinates (Xs
, Xe) and the Y coordinate (Y) are obtained, and step S110
3, the number of pixels of the n lines in the X direction (pixel) is obtained, and in step S1104, the number of pixels of the n lines (pi).
xel) is added to the number of pixels before the nth line to obtain the total number of pixels (area), and grv is calculated in step S1105.
Is calculated by the following equation (3). grv = grv + (pixel * Y) Equation (3) This grv can be grasped as an area moment (first moment) about the X axis by the pixel.

Thereafter, in step S1106, it is judged whether or not the calculation of the grv of the last line is completed. If the calculation of the grv of the last line is not completed yet, the calculation of the line in the X direction is completed in step S1107. The line number n is incremented, the process returns to step S1102, and the above process is repeated for the next unprocessed line. Then, after the calculation of the grv of the last line is completed, step S1
Proceeding to 108, the Y coordinate of the center of gravity is obtained by the following equation (4). gravY = grv / area (4) Expression

Here, the Y coordinate of the center of gravity of the binary image of FIG. 9 is calculated by the above method. First, in the processing of the first line, since there are no pixels and the number of pixels is 0, grv is also 0. In the processing of the second line, Xs = 1, Xe = 2, Y = 1,
Since pixel = 2 and area = 2, grv = 0
+ 2 * 1 = 2. In the processing of the third line, Xs = 1, Xe = 2, Y = 2,
Since pixel = 2 and area = 4, grv = 2
It becomes + 2 * 2 = 6. In the processing of the fourth line, Xs = 1, Xe = 4, Y = 3,
Since pixel = 4 and area = 8, grv = 6
It becomes + 4 * 3 = 18. In the processing of the fifth line, Xs = 1, Xe = 4, Y = 4,
Since pixel = 4 and area = 12, grv =
18 + 4 * 4 = 34.

With this, the processing for all the lines is completed, and the result is used to perform the processing for obtaining the Y coordinate gravY of the center of gravity. That is, gravY = 34/12 =
2.833 ... and the Y coordinate g of the center of gravity of the binary image in FIG.
It becomes rvY2.8. Therefore, the barycentric coordinates (gravX, gravY) of the binary image in FIG. 9 are (2.2,
2.8).

Next, a method of calculating the inclination will be described. FIG. 12 is a flowchart showing a process of obtaining a tilt by the conventional image processing apparatus.

In FIG. 12, X is determined in step S1201.
Initialize the line number n in the direction. Then, in step S1202, the cross-sectional synergistic moment (KnXY) of the pixel of the n line in the X direction is obtained, and in step S1203, the X
The second moment of area (InX) with respect to the X-axis of the pixel of the n-line in the X direction is obtained, and in Step S1204, the n in the X direction is calculated.
The second moment of area (In
Y) is asked. After this, in step S1205, it is determined whether or not the calculation for the last line has been completed. If the calculation for the last line has not yet been completed, step S1
In 206, the line number n of the line in the X direction is incremented, the process returns to step S1202, and the above process is repeated for the next unprocessed line. Then, after the calculation of the last line is completed, the process advances to step S1207 to calculate the barycentric coordinates (gravX, gravY). In addition,
The calculation of the barycentric coordinates can be obtained by the processing of FIGS. 10 and 11. In step S1208, the total number of pixels is added to obtain a total pixel area (area).
In step S209, the geometrical moment of inertia (KXY) of all pixels is calculated, in step S1210 the moment of inertia of area (IX) of all pixels with respect to the X-axis, and in step S1211, the moment of inertia of area (IY) of all pixels with respect to the Y-axis is calculated. The inclination (θ) is calculated in step S1212. This inclination (θ) is calculated by the following known equation (5). θ = (1/2) tan ^-1 {2KXY / (IY-IX)} Equation (5) In Equation (5), KXY is the cross-sectional synergistic moment of all pixels and IX is the X-axis of all pixels. Is the moment of inertia of area, IY is the moment of inertia of area for the Y axis of all pixels.

Here, KnXY, InX, and InY are obtained for each line of the binary image of FIG. 9 by the above method, and the inclination θ is calculated using this. First, in the processing of the first line, there are no pixels, and K1XY, I1X, and I1Y are all 0. In the processing of the second line, K2XY = 1 × 1 + 1 × 2 = 3,
I2X = 1 ² +1 ² = 2 and I2Y = 1 ² +2 ² = 5. In the processing of the third line, K3XY = 2x1 + 2x2 = 6,
I3X = 2 ² +2 ² = 8 and I3Y = 1 ² +2 ² = 5. In the processing of the 4th line, K4XY = 3 x 1 + 3 x 2 + 3 x
3 + 3 × 4 = 30, I4X = 3 ² +3 ² +3 ² +3 ² = 3
6, I4Y = 1 ² +2 ² +3 ² +4 ² = 30. In the processing of the 5th line, K5XY = 4 x 1 + 4 x 2 + 4 x
3 + 4 × 4 = 40, I5X = 4 ² +4 ² +4 ² +4 ² = 6
4, I5Y = 1 ² +2 ² +3 ² +4 ² = 30.

Area (area) = 12, barycentric coordinates (gravX, gravY) = (2.2, 2.
8) Therefore, the cross-sectional synergistic moment (KXY) of all pixels (KXY), X
The geometrical moment of inertia (IX) with respect to the axis and the geometrical moment of inertia (IY) with respect to the Y axis are as follows. KXY = (K2XY + K3XY + K4XY + K5XY)-(area x gravX x gravY) = (3 + 6 + 30 + 40)-(12 x 2.2 x 2.8) = 5.08 IX = (I2X + I3X + I4X + I5X)-(xrea x gra xra) (2 + 8 + 36 + 64)-(12 * 2.2 * 2.2) = 51.92 IY = (I2Y + I3Y + I4Y + I5Y)-(area * gravY * gravY) = (5 + 5 + 30 + 30)-(12 * 2.8 * 2.8) =- 24.08

Therefore, the inclination (θ) is θ = (1/2) tan ⁻¹ {(2 × 5.08) / (− 24.08−51.92)} −3.81 (degrees). .. In this way, the inclination (θ) of the binary image in FIG. 9 is obtained.

As another conventional image processing apparatus, the technique disclosed in Japanese Patent Laid-Open No. 2-250181 can be mentioned. This is to find the interior angle of each vertex of the quadrangle image obtained by the binarization process, and select a desired calculation procedure from the positional relationship in the quadrangle of two vertices having angles close to right angles among these angles. This is to obtain the rotational deviation angle.

[0018]

In the conventional image processing apparatus as described above, one CCD camera or the like captures an entire image of a quadrangular work, and the obtained image data is used to obtain the image shown in FIG.
The position and the inclination of the quadrangle work have been obtained by following the processing procedure from 0 to FIG. Therefore, it was not possible to detect the position and inclination of a large quadrangle work that does not fit within the field of view of the CCD camera.

Further, although the quadrangle work fits in the field of view of the CCD camera, the output result may not be obtained accurately due to the resolution of the image. For example, the resolution of an image is generally 256 × 256 pixels, 512 × 512 pixels, and if the field of view is fixed, it is possible to find how many millimeters each pixel corresponds to, and the number of millimeters per unit pixel is Since it is determined by the resolution of the image, it was not possible to expect a higher precision of the output result.

Therefore, an object of the present invention is to provide an image processing apparatus capable of detecting the position and inclination of a large quadrangle work which does not fit in the field of view of a CCD camera or a quadrangle work which requires high accuracy.

[0021]

An image processing apparatus according to the present invention processes image input means for capturing images of two vertices of a quadrangle work, and image data of two vertices captured by the image input means. And a position / inclination detecting means for detecting the position and inclination of the quadrangular work.

[0022]

In the image processing apparatus of the present invention, the image input means captures the images of the two vertices of the quadrangle work, respectively.
Since the position coordinates of the vertices are detected and the position and tilt of the quadrangle work are detected from the position coordinates, the position and tilt of a large quadrangle work that does not fit in the visual field of the image input means can be detected, and the image of the image can be detected. It is possible to detect the position and inclination of a quadrangular work with high accuracy without being affected by the resolution.

[0023]

EXAMPLES Examples of the present invention will be described below.

FIG. 1 is a flow chart showing a flow of processing units and operations of an image processing apparatus according to an embodiment of the present invention.

In the figure, a step S101 is a first image input section for performing an image input process, in which an image of the apex of a quadrangle work is picked up by an image pickup means such as a CCD camera to obtain its image signal, and this image signal is further obtained. Is A / D converted and output. Step S102 is a binarization unit that performs binarization processing, and the data digitized in step S101 is binarized to obtain a binary image of a rectangular work. In addition, the run length address data creation unit in step S103 regards the binarized pixel block of the rectangular work as the run length address data, and calculates the start and end addresses thereof. The circumscribed quadrangle extraction unit in step S104 obtains the upper left and lower right coordinates of the circumscribed quadrangle of the binary image based on the run length address data. Step S105 is a vertex calculation unit that calculates the coordinates of the vertices of the quadrangle work, and calculates the coordinates of the vertices of the quadrangle work from the two coordinates of the circumscribed quadrangle obtained by the circumscribed quadrangle extraction unit of step S104. And step S1
In 06, it is determined whether the apex is calculated by the image pickup means such as the first CCD camera or the like, or by the image pickup means such as the second CCD camera or the like, and the image pickup by the first CCD camera or the like is performed. If the vertices are calculated by the means, step S1
In step 07, the coordinates (X1, Y1) of the apex calculated by the image pickup means such as the first CCD camera are stored in the memory, and the process proceeds to step S108. Step S108 is Step S1
A second image input unit that performs the same image input processing as that of 01. The second apex of the quadrangle work is imaged by an image pickup means such as a second CCD camera, and its video signal is obtained.
Further, this video signal is A / D converted and output. After this, the process returns to step S102 again, and the first CC
A processing procedure similar to that of the image pickup means such as a D camera is performed. On the other hand, when the vertices are calculated by the image pickup means such as the second CCD camera in step S106, the coordinates (X2, Y2) of the vertices calculated by the image pickup means such as the second CCD camera in step S109 are stored in the memory. save. Step S11
At 0, the coordinates (X1, Y1), (X2 of the vertices calculated from the image pickup means such as the CCD cameras of the first and second units, respectively.
, Y2), the coordinates of the center and the tilt angle are obtained. However, the circumscribed rectangle is the coordinates of the two vertices, that is, P1 (Xmi
n, Ymin) and P2 (Xmax, Ymax) are output.

Of the respective elements in FIG. 1 constituting the present embodiment, the first image input section in step S101, the binarization section in step S102, the run length address data creation section in step S103, and step S104. The details of the circumscribed quadrangle extraction unit and the second image input unit of step S108 have already been proposed and are publicly known or can be easily inferred from digital processing, and thus description thereof will be omitted. Here, the vertex calculating unit (step S105), which is the largest feature of this embodiment,
The process (step S110) of detecting the position and inclination of the quadrangle work from the coordinates of one vertex will be specifically described.

First, an outline of image input of a rectangular work will be described with reference to FIG. FIG. 2 is an explanatory diagram showing an outline of image input of a square work by the image processing apparatus according to the embodiment of the present invention.

In the figure, 1 is a square work, 2 is a field of view of a first CCD camera, 3 is a field of view of a second CCD camera, and 4 is a square work 1 by a first CCD camera.
Is a circumscribing quadrangle circumscribing one of the vertices of the same, and 5 is a circumscribing quadrangle circumscribing the other apex of the quadrangle work 1 by the second CCD camera. As shown in FIG. 2, the two vertices of the slightly inclined quadrangle work 1 are set so as to be in the fields of view 2 and 3 of the first and second CCD cameras, respectively. Perform value conversion processing.

FIG. 3 shows a binarization process after image input by a CCD camera of an image processing apparatus according to an embodiment of the present invention.
It is explanatory drawing which shows the state which calculated | required the circumscribed quadrangle. In this figure, it can be assumed that the desired vertex corresponds to (0,5) or (0,6).

Next, referring to FIG.
Starting from FIG. FIG. 4 is an explanatory diagram showing an image input state of a square work by the image processing apparatus according to the embodiment of the present invention.

In the figure, the state (a) is the first C in FIG.
The image of the visual field 2 of the CD camera is shown, and the shaded area shows the binary image of the quadrangle work 1. The image in this state (a) shows that the square work 1 is tilted clockwise. The dotted line represents the circumscribed quadrangle of the binary image. The point 1 corresponds to the desired vertex of the square work 1.
At this time, attention is paid to the pixels existing in the corner 1 and the corner 2. Corner 1 of state (a), that is, the surrounding three pixels (0, 0) having the coordinate value (0, 0) in FIG.
No pixel exists in any of 1), (1, 0), and (1, 1). In addition, the corner 2, that is, the surrounding 3 pixels (8, 7), (7, 7) having the coordinate value (8, 8) in FIG.
Pixels are present in any of 8) and (7, 7) (in FIG. 3, pixels are present in all). The condition (a) is defined when such a condition is satisfied.

The state (b) is the same as the state (a),
The image of the field of view 2 of the first CCD camera of FIG. 2 is shown, and the image in this state (b) shows that the square work 1 is tilted counterclockwise. Corner 1 and corner 2
Attention is paid to the pixels existing in 1), there is a pixel in any of the 4 pixels in the corner 1, and there is no pixel in any of the 4 pixels in the corner 2. The condition (b) is defined when such a condition is satisfied.

Similarly to the states (a) and (b), the state (c) also shows an image of the visual field 2 of the first CCD camera shown in FIG. Focusing on the pixels existing in the corner 1 and the corner 2, the pixel exists in any of the four pixels of the corner 1 and the corner 2. The condition (c) is defined when such a condition is satisfied.

As described above, the image pattern of the field of view of the CCD camera is roughly divided into three states, and the processing corresponding to each state is performed. FIG. 5 shows the flow of processing divided into these three types of states. FIG. 5 is a flow chart showing the operation of distributing the image input to three kinds of states by the image processing apparatus according to the embodiment of the present invention. In this figure, the binarization process, the run length address data creation process, and the circumscribed quadrangle extraction process have already been completed after the image is input, and the program for executing the process of FIG. 5 is executing the main program of the image processing. Be called by.

In the figure, first, in step S501, it is determined whether or not a pixel exists in the corner 1.
If the pixel exists in the corner 1, step S502
Then, it is determined whether or not a pixel exists in the corner 2.
If the pixel exists in the corner 2, the pixel exists in both the corner 1 and the corner 2, so that the step S
Proceeding to 503, the processing of state (c) is performed. If there is no pixel in the corner 2 in step S502,
The state (b) is processed in step S504. On the other hand, if the pixel does not exist in the corner 1 in step S501, it is determined in step S505 whether or not the pixel exists in the corner 2. If the pixel exists in the corner 2, the state (a) is processed in step S506. If the pixel does not exist in the corner 2 in step S505, there is no pixel in the corner 1 and the corner 2. In step S507, error processing is performed. In this way, the states of the pixels at the corner 1 and the corner 2 of the circumscribed quadrangle are determined, and the state (a) is determined according to the state.
To the state (c), and the vertex calculation processing operation is performed according to each of the three types of states.

Here, the vertex calculation processing operation corresponding to each of the above-mentioned three types of states will be described. FIGS. 6 to 8 are flowcharts corresponding to the respective state processes of the states (a) to (c) of FIGS. 4 and 5, and FIG. 6 is the vertex calculation processing operation of the states (a) of FIGS. 4 and 5. FIG. 7 shows FIG. 4 and FIG.
8 shows the vertex calculation processing operation in the state (b), and FIG. 8 shows the vertex calculation processing operation in the state (c) of FIGS. 4 and 5.

First, in FIG. 6, in step S601, Xmin of the circumscribing quadrangle is fixed, one line on the Y axis is searched in the Y direction to check the presence or absence of each pixel, and step S60 is performed.
The address of the pixel existing in 2 in the Y direction is stored. Then, in step S603, it is determined whether or not there is only one pixel in the Y direction. If there is only one pixel, then in step S604, the address of the pixel, that is, (Xmin, Y) is set as the vertex coordinate. If there are a plurality of pixels in the Y direction, the average value of the maximum Ymax and minimum Ymin address values is calculated in step S605, and the average value is adopted in step S606 to obtain (Xmin, (Ymax + Ymin ) / 2) is the vertex coordinate. In this way, the vertex of the square work 1 in the state (a) of FIG. 4 is obtained.

Further, in FIG. 7, Ymax of the circumscribing quadrangle is fixed in step S701, a search is made on one line of the X axis in the X direction to check the presence or absence of each pixel, and step S70
The address of the pixel existing in 2 in the X direction is stored. Then, in step S703, it is determined whether or not there is only one pixel in the X direction. If there is only one pixel, in step S704, the address of the pixel, that is, (X, Ymax) is set as the coordinate of the vertex. If there are a plurality of pixels in the X direction, the average value of the maximum Xmax and the minimum Xmin is calculated in step S705, and the average value is adopted in step S706.
Let ((Xmax + Xmin) / 2, Ymax) be the coordinates of the vertex. In this way, the vertex of the square work 1 in the state (b) of FIG. 4 is obtained.

Then, in FIG. 8, step S801.
Then, the circumscribed quadrangle (Xmin, Ymax) is unconditionally used as the coordinate of the vertex. In this way, the vertex of the square work 1 in the state (c) of FIG. 4 is obtained.

After the image input processing by the first CCD camera is completed as described above, the image input processing by the second CCD camera is similarly performed, and the coordinates of the two vertices of the square work 1 are determined. Ask each. Then, the center coordinates and the inclination of the square work 1 are detected from the coordinates of these two vertices. Since the detection of the center position coordinates and the inclination can be easily obtained by a mathematical method, the description thereof will be omitted here.

As described above, the image processing apparatus of this embodiment is
Image input means composed of first and second two CCD cameras for respectively capturing images of two vertices of the square work 1, and image data of two vertices captured by the two CCD cameras (image input means). Based on the above, the coordinates of the two vertices are obtained by the binarization process, the run length address creation process, the circumscribing quadrilateral extraction process, and the vertex calculation process of FIGS. And a position / tilt detecting means for detecting the position and the tilt.

That is, the image processing apparatus of the present embodiment captures images of two vertices of the quadrangle work 1 with two CCD cameras, detects the position coordinates of these two vertices, and detects the position coordinates of the quadrangle work 1 from the position coordinates. The position and the inclination are detected.

Therefore, it is not necessary to take the whole image of the square work 1 with one CCD camera and to obtain the position and inclination of the square work using the obtained image data as in the conventional case. It is possible to detect the position and inclination of the large quadrangle work 1 that does not fit in the field of view. Further, by appropriately bringing the two CCD cameras close to the two vertices of the square work 1, the number of millimeters per unit pixel can be adjusted regardless of the resolution of the image of the CCD camera, so that the accuracy of the output result is improved. be able to. Therefore, the position and inclination of the square work 1 can be accurately detected without being affected by the image resolution. As a result, the applicable range is expanded, and the position and inclination of the quadrangle work 1 that requires high accuracy can be detected.

By the way, in the above embodiment, one of the two CCD cameras for inputting the images of the two vertices of the square work is fixed, the other one is attached to the XY table or the like, and the CCD camera is moved to the image input position. The same effect can be obtained by moving the image, inputting images of two vertices, and performing the process.

Further, in the above-mentioned embodiment, the square work 2 is used.
Although the image input of the vertices was performed at diagonal corners, the same effect can be obtained by selecting any two vertices of the quadrangle work, performing the image input on the vertices, and performing the processing. ..

Further, by using the coordinates of the two vertices of the quadrangle work, it is possible to calculate not only the position and inclination, but also many feature quantities such as area and perimeter.

[0047]

As described above, the image processing apparatus of the present invention comprises image input means and position / tilt detection means,
By capturing the images of the two vertices of the quadrangle work by the image input means, detecting the position coordinates of these two vertices, and detecting the position and inclination of the quadrangle work from these position coordinates, the image does not fall within the visual field of the image input means. It is possible to detect the position and tilt of a large quadrangle work, and it is possible to detect the position and tilt of the quadrangle work with high accuracy without being affected by the resolution of the image. And the inclination can be detected.

[Brief description of drawings]

FIG. 1 is a flowchart showing a flow of processing units and operations of an image processing apparatus which is an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an outline of image input of a square work by an image processing apparatus according to an embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a state in which a circumscribed quadrangle is obtained by performing binarization processing after image input by the image processing apparatus according to the embodiment of the present invention.

FIG. 4 is an explanatory diagram showing an image input state of a square work by the image processing apparatus which is an embodiment of the present invention.

FIG. 5 is a flowchart showing an operation of distributing an image input to three types of states by the image processing apparatus according to the embodiment of the present invention.

FIG. 6 is a flowchart showing the vertex calculation processing operation in the state (a) of FIGS. 4 and 5;

FIG. 7 is a flowchart showing the vertex calculation processing operation in the state (b) of FIGS. 4 and 5;

FIG. 8 is a flowchart showing the vertex calculation processing operation in the state (c) of FIGS. 4 and 5;

FIG. 9 is an explanatory diagram showing a binary image by a conventional image processing apparatus.

FIG. 10 is an X of the center of gravity by the conventional image processing apparatus.
It is a flowchart which shows the process which calculates | requires a coordinate.

FIG. 11 is a graph showing Y of the center of gravity by the conventional image processing apparatus.
It is a flowchart which shows the process which calculates | requires a coordinate.

FIG. 12 is a flowchart showing a process of obtaining a tilt by a conventional image processing apparatus.

[Explanation of symbols]

 1 square work 2 field of view of first CCD camera 3 field of view of second CCD camera 4 circumscribed quadrangle 5 circumscribed quadrangle

[Procedure amendment]

[Submission date] December 11, 1991

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0005

[Correction method] Change

[Correction content]

First, a method of calculating the X coordinate will be described. In FIG. 10, in step S1001, the line number n in the X direction is initialized. Then, step S10
At 02, the start and end coordinates (X
s, Xe), and n in the X direction is calculated in step S1003.
The number of pixels of the line (pixel) is calculated, and step S10
In 04, the number of pixels of the n line (pixel) is added to the number of pixels before the n line to obtain the total number of pixels (area). In step S1005, grv is calculated by the following equation (1). grv = grv + {pixel × (Xs + Xe)} / 2 (1) This grv can be understood as an area moment (first moment) of the pixel about the Y axis.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0007

[Correction method] Change

[Correction content]

Here, the X coordinate of the center of gravity of the binary image of FIG. 9 is calculated by the above method. First, in the processing of the first line, since there are no pixels and the number of pixels is 0, grv is also 0. In the processing of the second line, Xs = 1, Xe = 2, pixel
Since l = 2 and area = 2, grv = 0 + {2 ×
(1 + 2)} / 2 = 3. In the processing of the third line, Xs = 1, Xe = 2, pixel
Since l = 2 and area = 4, grv = 3 + {2 ×
(1 + 2)} / 2 = 6. In the processing of the fourth line, Xs = 1, Xe = 4, pixel
Since l = 4 and area = 8, grv = 6 + {4 ×
(1 + 4)} / 2 = 16. In the processing of the fifth line, Xs = 1, Xe = 4, pixel
Since l = 4 and area = 12, grv = 16 +
{4 × (1 + 4)} / 2 = 26. This completes the processing for all the lines, and the processing for obtaining the X-coordinate gravX of the center of gravity is performed using this result. That is, gravX = 26/12 = 2.166 ..., and the X coordinate of the center of gravity of the binary image in FIG. 9 gravX = 2.
It becomes 2.

[Procedure amendment 3]

[Document name to be amended] Statement

[Correction target item name] 0008

[Correction method] Change

[Correction content]

Next, a method of calculating the Y coordinate will be described. In FIG. 11, in step S1101, the line number n in the X direction is initialized. Then, step S11
02, the start and end coordinates (Xs
, Xe) and the Y coordinate (Y) are obtained, and step S110
3, the number of pixels of the n lines in the X direction (pixel) is obtained, and in step S1104, the number of pixels of the n lines (pi).
xel) is added to the number of pixels before the nth line to obtain the total number of pixels (area), and grv is calculated in step S1105.
Is calculated by the following equation (3). grv = grv + (pixel × Y) Equation (3) Note that this grv can be understood as an area moment (first moment) of the pixel about the X axis.

[Procedure correction 4]

[Document name to be amended] Statement

[Correction target item name] 0010

[Correction method] Change

[Correction content]

Here, the Y coordinate of the center of gravity of the binary image of FIG. 9 is calculated by the above method. First, in the processing of the first line, since there are no pixels and the number of pixels is 0, grv is also 0. In the processing of the second line, Xs = 1, Xe = 2, Y = 1,
Since pixel = 2 and area = 2, grv = 0
+ 2 × 1 = 2. In the processing of the third line, Xs = 1, Xe = 2, Y = 2,
Since pixel = 2 and area = 4, grv = 2
+ 2 × 2 = 6. In the processing of the fourth line, Xs = 1, Xe = 4, Y = 3,
Since pixel = 4 and area = 8, grv = 6
+ 4 × 3 = 18. In the processing of the fifth line, Xs = 1, Xe = 4, Y = 4,
Since pixel = 4 and area = 12, grv =
18 + 4 × 4 = 34.

[Procedure Amendment 5]

[Document name to be amended] Statement

[Correction target item name] 0011

[Correction method] Change

[Correction content]

With this, the processing for all the lines is completed, and the result is used to perform the processing for obtaining the Y coordinate gravY of the center of gravity. That is, gravY = 34/12 =
2.833 ... and the Y coordinate g of the center of gravity of the binary image in FIG.
rabY = 2.8. Therefore, the barycentric coordinates (gravX, gravY) of the binary image in FIG. 9 are (2.2,
2.8).

[Procedure Amendment 6]

[Document name to be amended] Statement

[Correction target item name] 0016

[Correction method] Change

[Correction content]

Therefore, the slope (θ) is θ = (1/2) tan ^-1 {(2 × 5.08) / (-2
4.08-5.92)} ≈ −3.81 (degrees).
In this way, the inclination (θ) of the binary image in FIG. 9 is obtained.

[Procedure Amendment 7]

[Document name to be corrected] Drawing

[Correction target item name] Figure 9

[Correction method] Change

[Correction content]

[Figure 9]

[Procedure Amendment 8]

[Document name to be corrected] Drawing

[Name of item to be corrected] Fig. 10

[Correction method] Change

[Correction content]

[Figure 10]

[Procedure Amendment 9]

[Document name to be corrected] Drawing

[Name of item to be corrected] Fig. 11

[Correction method] Change

[Correction content]

FIG. 11

Claims (1)

Claim: What is claimed is: 1. An image input unit for capturing images of two vertices of a quadrangle work, and image data of the two vertices captured by the image input unit for processing the position and position of the quadrangle work. An image processing apparatus comprising: a position / tilt detecting means for detecting an inclination.