JPH0511046A - Method and apparatus for detecting position of body provided in concealed place - Google Patents

Abstract

PURPOSE:To make it possible to find the accurate position of a body based on only the image of the identified body by forming point-shaped images and binary-coding the images by a plurality of the different means, obtaining the point parts and the maximum gradations within a common region wherein the points are overlapped, and selecting the parts and the gradations in the sequence of the magnitudes. CONSTITUTION:The electromagnetic waves of pulse signals are emitted from an oscillating means comprising an oscillating circuit 4, a transmitting circuit 5 and an antenna 6 on a pipe 2 which is embedded in the soil. The reflected waves from the pipe 2 are received with an antenna 7 and imparted into a receiving circuit 8. The original image of the cross section including the pipe 2 is obtained by fixing the antennas 6 and 7 as a unitary body and by the movement crossing the pipe 2 as shown with an arrow 9. The output of the circuit 8 is imparted into a processing circuit 10 of a microcomputer. In the circuit 10, the original image is stored in a memory 11. The original image is read, and point-shaped images are formed by a plurality of means and binary-coded. The point parts within a common region wherein the points of the binary-coded image-shaped image are overlapped are obtained based on the number of the detected bodies as the inputs of an input means 60. The results of the operation of the maximum gradations are selected in the sequence of the magnitudes. The images are displayed on a display device 12 based on the result so that the image can be identified.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for detecting an object provided in a concealed place, such as an underground pipe, and detecting the position of the object.

[0002]

2. Description of the Related Art A ground-penetrating radar for detecting a pipe buried in the ground that transports gas etc. on the ground emits a pulsed electromagnetic wave into the ground and reflects a wave reflected by the underground pipe. An image of the cross section of the soil containing the buried pipe can be obtained by receiving at the antenna and thus scanning the buried pipe across the ground. The images obtained by such prior art are disturbed by strong reflection images on the ground surface, and also by virtual images due to underground water in the ground, cavities and foreign soil boundaries, and also by electrical noise. Therefore, the reflection image of the underground buried pipe, which is the target object to be searched, is rarely seen clearly.

In order to solve such a problem, an attempt is made to improve the image quality of the original image by improving the ground survey radar.
And image processing is performed by signal / image processing to improve image quality, and feature extraction is performed. Such prior art is known as difference processing, product processing, migration, deconvolution, synthetic aperture, and the like, and examples thereof include JP-A-63-222284. Even in such respective prior arts, it is difficult to reliably detect the detection of the underground buried pipe.

In the image measured by the above-mentioned underground survey radar, the reflected wave of the pipe becomes an arc. Based on this, the position where the pipe is buried is judged by human eyes. But stones in the soil,
There are cavities, groundwater, geological changes, and other buried objects, and part of the arc disappears and merges with other reflection images, which greatly influences the interpretation. For this reason, it is necessary to have the knowledge and experience of a skilled person to read the image. However, it takes a long time to train skilled workers. This hinders the spread of radar locators. In other words, the image obtained by this underground survey radar requires the knowledge and experience of a skilled person. There are also oversights of experts.

[0005]

SUMMARY OF THE INVENTION The present invention is capable of detecting an object provided in a hidden place such as the underground, that is, detecting the position of the object, even if the person is unfamiliar, unfamiliar, or layman. The object of the present invention is to provide a method and apparatus for detecting the position of an object provided in a concealed place that can be performed.

[0006]

According to the present invention, an original image of an object provided at a concealed place is obtained, and a dot-like image having a plurality of gradations is obtained by a plurality of different methods, and each dot-like image is obtained. The image is binarized, and the point portion in the common area where the respective points of the binarized point-shaped image overlap is obtained, and for each point-shaped image, the maximum gradation in each point portion and its maximum Gradient coordinates are calculated, the maximum gradation of the point part in each common area is mutually calculated, the number of objects to be detected is input, and the calculation result for each common area is calculated in order of magnitude. In the concealed place, which is characterized in that the input number is selected, and only the image of the object corresponding to the selected common area is identifiable on the original image, and is superimposed and displayed based on the coordinates. This is a method for detecting the position of an object provided.

Further, the present invention is characterized in that each point of the binarized point-shaped image is expanded, then the point portion in the common area of the points is obtained, and this point portion is expanded. .

The present invention also provides (a) one of the methods described above,
(A1) Each pixel of the original image and another one with a predetermined fixed interval in the direction perpendicular to the first line of symmetry with respect to the pixel
The level difference of the signal with two pixels is obtained, the level difference is stored for each pixel, the difference processing is performed, and the secondary image is obtained,
(A2) Based on the secondary image, the signal levels of pixels on a predetermined line that is line-symmetric with respect to the second symmetry line parallel to the first symmetry line are added on the left and right sides of the second symmetry line, and The difference ΔE between the sum E1 and the sum E2 on the right side is calculated,
The difference ΔE is stored as a value on the pixel at the apex of the line, and (a3) the second symmetry line is shifted in the direction perpendicular to the first symmetry line in a state parallel to the first symmetry line to obtain the difference ΔE. The operation of obtaining and storing as the value on the pixel of the vertex is repeatedly performed over the entire image to obtain the dot-shaped image, and (b) the other method is
(B1) A second line parallel to the first line of symmetry based on the original image
The signal levels of pixels on a predetermined line that is line-symmetric with respect to the line of symmetry are added on the left and right sides of the second line of symmetry,
The sum E of the left-side sum E1 and the right-side sum E2 is obtained, and the sum E is stored as a value on the pixel at the apex of the line,
(B2) The first line of symmetry is parallel to the first line of symmetry
It is characterized in that a dot-like image is obtained by repeating the operation of shifting the symmetry line in the vertical direction to obtain the sum E and storing it as a value on the pixel at the apex over the entire image.

Further, according to the present invention, an oscillating means 4, 5 and 6 for radiating a pulse signal and a hiding place including an object to be detected by receiving a reflected wave by the oscillating means and obtaining a time difference W1 from emission to reception. Is obtained and stored in the memory 11, and means 7 to 12 for visually displaying by the display means 12 and the original image stored in the memory are read out and read by a plurality of different methods. A dot-shaped image creating unit that obtains and creates a dot-shaped image, and binarizes each dot-shaped image in response to the output from the dot-shaped image creating unit 2
Quantifying means and input means 6 for inputting the number of objects to be detected
0, in response to each output of the binarizing means and the input means 60,
The point portion in the common area where the respective points of the binarized point-shaped image overlap is obtained, and the maximum gradation in each point portion and the coordinates of the maximum gradation are obtained for each point-shaped image. A calculating means for calculating the maximum gradations of the dot portions in each common area, and selecting the calculation result for each common area in the order of magnitude of the calculation result by the input number; In response to the output of the display means 12, a means for superimposing and displaying an identification symbol on the original image displayed by the display means 12 in the vicinity of the image of the object corresponding to the selected common area. A position detecting device for an object provided in a concealed place, which includes:

[0010]

According to the present invention, first, an original image of an object provided at a concealed place, such as an underground buried pipe, is obtained. With respect to this original image, for example, a pulse signal is radiated, a reflected wave from an object provided in the hiding place is received, the time difference W1 from the emission to the reception is obtained, and the original image of the hiding place including the object to be detected Can be obtained. This image processing of the original image is performed to form a dotted image by a plurality of different methods. Each dot-shaped image has a plurality of gradations. The punctiform image is a candidate point of the position where the object exists, which is obtained from the original image. Each point of this punctiform image includes not only an accurate representation of the position of the object but also a representation of a point due to disturbance, even though it is not an object.

As another embodiment, in obtaining an original image, each of the original images is obtained by a plurality of different methods, and based on each original image, for example, the same image processing method has a plurality of gradations. You may make it obtain each dot-shaped image.

Each dot-shaped image thus obtained is binarized, and a point portion in a common area where each point of the binarized dot-shaped image overlaps is obtained. Next, based on each dot-shaped image, the maximum gradation in each point portion and the coordinates of the pixel having the maximum gradation in, for example, the x, y coordinate system are obtained. The maximum gradations thus obtained in the dot portions for each dot-shaped pixel in each common area are mutually calculated, for example, multiplied or added. In this way, the higher the reliability of the candidate point of the position of the object obtained by the pointized image obtained by each of the plurality of different methods, the larger the value of the calculation result obtained by mutually calculating the maximum gradation. The operator inputs the number of objects existing in the range of the hiding place detected by the original image by using, for example, a key input means. The number of objects may be prepared and stored in a memory or the like in advance. Therefore, the calculation results for each common area are arranged in the order of magnitude of the calculation results, that is, from the largest to the smallest, and the selected number is selected in order from the largest calculation result. Therefore, only the image of the object on the original image is displayed in a distinguishable manner. In order to be able to identify only the image of this object, an identification symbol, for example, near the image of the object on the original image, ie on the image of the object or at a position slightly distant from the image of the object, It may be marked with a circle, or may be displayed in a distinguishable manner by blinking or changing the display color.

Furthermore, according to the present invention, each point of the binarized point-shaped image is expanded, and after obtaining the point portion, the point portion is expanded so that an object may exist. It is possible to catch a candidate point without missing it.

Further, according to the present invention, as one method for obtaining the dot-shaped image, another pixel is provided in the original image with a predetermined constant interval in the direction perpendicular to the first symmetry line with respect to the pixel. The level difference of the signal with the image is obtained, and the level difference is stored for each pixel. In this way, the difference processing is performed,
Find the secondary image. Such difference processing has a function of weakening a uniform image continuous in the direction perpendicular to the first symmetry line, for example, in the horizontal direction. Therefore, for example, it is possible to suppress the obstruction of images such as groundwater horizontally existing in the ground.

Next, based on this secondary image, the signal levels of pixels on a predetermined line symmetric with respect to the second symmetry line, for example, a circle or a hyperbola, are added on the left side of the second symmetry line, and the sum E1 is obtained. And add on the right side to get the sum E2
Then, the difference ΔE between the sums E1 and E2 is obtained, and the difference ΔE thus obtained is stored as the value on the pixel at the apex on the second line of symmetry of the line. This operation is performed in the first state with the second line of symmetry parallel to the first line of symmetry.
It is repeated while shifting in the direction perpendicular to the line of symmetry to obtain a dot-like image. According to the so-called inverted synthetic aperture processing, the feature extraction of the vertex of the object can be clearly performed without being adversely affected by the image extending in the vertical direction of the first and second symmetry lines, for example, in the horizontal direction. Like According to the present invention, the difference processing is performed based on the original image as described above.
The next image is obtained, and then the inversion synthetic aperture process is performed to obtain the dot image.

As another method for obtaining the dot-shaped image, instead of obtaining the above-mentioned difference ΔE for the image by performing the difference processing on the original image, the sum E1 for the original image is obtained.
The sum E of E2 may be obtained, and the other procedure is the same as the above-described method.

[0017]

1 is a block diagram showing the electrical construction of an embodiment of the present invention. A pipe 2 such as a steel pipe for transporting a fluid such as gas is buried in the soil 1, and the pipe 2 is line-symmetrical to the left and right in FIG. In order to detect this tube 2 on the ground, the oscillation circuit 4 is shown in FIG.
For example, 100 MHz to 1000 MH shown in (1)
A pulse having a waveform of z is generated, and a transmission circuit 5 including the oscillation circuit 4 supplies an electric signal having a waveform shown in FIG. This makes the antenna 6
Radiates the electromagnetic wave shown in FIG. 2C to the soil 1. The electromagnetic wave radiated from the antenna 6 is reflected by the tube 2, and the reflected wave is received by the antenna 7 and given to the receiving circuit 8. The electromagnetic wave shown in FIG. 2C is radiated from the antenna 6 and the antenna 7
The time difference W1 until the signal shown in FIG. 2 (4) is received corresponds to the distance between the surface of the soil 1 and the pipe 2. By fixing the transmitting antenna 6 and the receiving antenna 7 integrally and moving them so as to traverse the underground buried pipe 2 as shown by the arrow 9 in FIG.
An original image of the cross section in the vertical plane of can be obtained.

The output of the receiving circuit 8 is given to a processing circuit 10 realized by a microcomputer or the like. A memory 11 is connected to the processing circuit 10, and the memory 1
In 1 is stored the original image, secondary images and dot images by the first and second methods described later. Visual display means 12 realized by a cathode ray tube or the like
, These images are visually displayed. The secondary image and the punctured image are obtained by two different methods in this embodiment. The key input means 60 is connected to the processing circuit 10.

FIG. 3 is a flow chart for explaining the operation of the processing circuit 10. The process moves from step c1 to step c2 of FIG. 3, and the memory 11 stores the original image shown in FIG. This original image is visually displayed by the display means 12. It is difficult for a person who has no knowledge, an unfamiliar person, or an amateur to clearly read and know where the tube 2 is buried just by looking at the original image of FIG. This is because the virtual image may appear in the original image. The picture elements of the original image are arranged in a matrix in the x, y rectangular coordinate system of the original image, and the coordinates of each pixel and the plurality of gradations of the pixel are shown in FIG. Stored as shown. In this embodiment, the gradation is 0 in total.
.About.255, and has a total of 256 gradations.

In step c2a, the first punctiform image is
The first punctiform image created based on the original image of FIG. 4 and created by this first method is as shown in FIG. At step c3, the first punctiform image shown in FIG. 6 is binarized. In this binarization, gradation zero is defined as logic "0", and gradations 1 to 255 are defined as logic "1". The binarized image of the first dot-shaped image thus obtained is as shown in FIG. 7. In FIG. 6, pixels with small gradation cannot be clearly seen, but in the binarized image of FIG. 7, pixels with gradation 1 or more can be clearly seen. The binarized image of FIG. 7 thus obtained is expanded in step c4 as shown in FIG. In order to expand the image, first expand the vicinity of 4,
Next, expansion in the vicinity of 8 is performed. As shown in FIG. 9A, when the pixel 41 has the logic “1”, the expansion in the vicinity of 4 means that the four pixels 42 to 45 adjacent to the upper, lower, left, and right sides of the coordinate position have the logic “1”. ". When a pixel 46 exists near the pixel 41 as shown in FIG. 9 (2) and all of them are logic "1", expansion of 4 neighborhoods is performed for this pixel 41 as indicated by reference numeral 47, Similarly, expansion of the pixel 46 in the vicinity of 4 is performed as indicated by reference numeral 48. As a result of the dilation of each pixel 41, 46, the pixels 49, 50 that overlap as a logic "1" remain at a logic "1".

After the 4-neighbor expansion of the pixel 41 of logic "1" in FIG. 9 (1) is performed in this manner, the 8-neighbor expansion is performed on each of the expanded pixels 42 to 45 as shown in FIG. Do each. For example, regarding the coordinate 42, a total of eight pixels surrounding the coordinate 42 are all set to the logic "1". Thus, for each pixel of the binarized image shown in FIG. 7, the 4-neighbor expansion is first performed as shown in FIG. 9, and further the 8-neighbor expansion is performed as shown in FIG. To do.

At step c4a, a second point-shaped image is obtained as shown in FIG. 11 by the second method described later, and at the next step c5, the second point-shaped image is binarized. A binarized image is created as shown in FIG.
The procedure for creating this binarized image is as described in step c3 above.
Is the same as. In step c6, the binarized image obtained as shown in FIG. 12 is subjected to 4-neighbor expansion and 8-neighbor expansion work in the same manner as in the above-mentioned step c4, and as shown in FIG. Obtain an image that has been expanded once.

In step c7, the dilated punctured image of FIG. 8 based on the first technique and FIG. 1 based on the second technique.
The point portion within the common region where the points of the dilated pointized image shown in FIG. 3 overlap each other is obtained, and the image of the point portion is obtained as shown in FIG. That is, as shown in FIG. 15, a dilated point 51 of FIG. 8 and a dilated point 52 of FIG. 13 are shaded so that a common area 53 is obtained. , As shown in FIG. In this way, AND processing of the points 51 and 52 is performed.

At step c8, the AN shown in FIG.
The image of FIG. 16 is obtained by performing the expansion calculation process of the D-processed image. This expansion calculation processing is performed in the vicinity of 4 and / or as described above.
Alternatively, the expansion may be around 8. Each inflated point is indicated by reference numeral 54.

Referring to FIG. 17, when the picture element of the first point-shaped image by the first method is 41, the point expanded in the vicinity of 4 is indicated by reference numeral 55, and further expanded in the vicinity of 8 by And get point 51. Pixel 5 of the second dot-shaped image by the second method
Inflate 6 near 4 to get point 57 and near 8 to get point 52. A common area 53 shown by hatching in FIG. 17 is expanded to obtain a point 54. The reason why the expansion of the image after the AND processing in step c8 is necessary is to refer to FIG. 18 and the pixel 41 of the first dot-shaped image and the pixel 56 of the corresponding second dot-shaped image. If and are relatively far apart,
When 1 and 56 are respectively expanded twice to obtain points 51 and 52, and the common area 53 is obtained by ANDing them, the common area 53 is obtained by the first and second methods. The pixels 41 and 56 which are the candidate points of the tube 2 are not included, which may make it impossible to detect the position of the tube 2 with high reliability. Therefore, in step c8, the common area 53, for example, in the embodiment of FIG. 18, the point 54 after the expansion is obtained by sequentially performing the 4-neighbor expansion and the 8-neighbor expansion. Thus, the common area 53
The points 54 obtained by inflating include the pixels 41 and 56 which are the candidate points of the tube 2, and thus the position of the tube 2 can be detected with high reliability.

In step c9, each point 54 of the above-described image of FIG. 16 consisting of dilated points 54 is numbered and labeled to uniquely identify it. That is, among the pixels included in the point 51 based on the first method, the coordinates (x, y) of the pixel included in the pixel 41 within the expanded point 54 and the respective gradations of the pixel are shown in FIG. ) Store in the memory 11 as shown in FIG. Similarly, as shown in FIG. 19 (2), among the points 52 based on the second method, the pixel 5 within the expanded point 54
The pixels included in 6 are numbered, and the coordinates (x, y) of each pixel and each gradation are stored in the memory 11.

Then, in step c10, the pixel 41
, The maximum value M1 of the gradations of those pixels, for example, the gradation 254 in FIG.
Ask for. Similarly, among the pixels included in the pixel 56, the maximum value M of the gradations of those pixels
2, for example, in FIG. 19B, the gradation 249 is obtained. The product M of the maximum values M1 and M2 of each pixel included in the pixels 41 and 56 thus obtained is calculated to obtain the product M.

[0028]

## EQU1 ## M = M1.M2 Further, the coordinates of the pixel having such a maximum gradation, for example, (x1, y1) in FIG. 19 (1), and FIG. 19 (2).
Then, the coordinates are (x2, y2), and the average coordinates (x12, y12) of these coordinates are obtained as shown in Equation 2.

[0029]

[Equation 2]

Such steps c9 and c
Ten calculations are performed for each inflated point 54. Such a point 54 can be considered as a window, so to speak, and the product M of the maximum gradations M1 and M2 of the pixels 41 and 56 viewed from this window.
Is calculated as described above, and the average coordinates (x12, y12) of the maximum values M1 and M2 are calculated.

At step c11, the search range in which the pipe 2 is embedded is entered from the key input means 60. For example, in the original image shown in FIG. 4, the horizontal distance W1 and the depth D1 are input from the key input means 60 and given to the processing circuit 10. With this, even if the original image includes a virtual image or the like, it is possible to reliably prevent erroneous determination as the tube 2 by mistake. Note that this step c11 may be omitted.

At step c12, the search range W1 × D1
The expected value of the number of displayed tubes 2 inside is input from the key input means 60. For example, in this embodiment, the expected number of display lines is 6.

In step c12a, the products M, which are the calculation results corresponding to the above-mentioned large number of inflated point portions 54, are arranged in order from the largest to the smallest, and the products are input by the number input in step c12. The one with the largest M is selected in order, and the average coordinates (x12, y
12) is obtained in step c12b.

In step c13, the image of FIG. 4 which is the original image is displayed by the display means 12, and on the original image, in step c14, the tube corresponding to the dilated point 54 of the selected common area 53 is displayed. So that only the two images can be identified, the coordinates (x12, y1) of the selected pixel are selected.
As shown in FIG. 20, for each 2), an identification symbol, for example, ○
The mark 61 is superimposed and displayed. Therefore, by looking at the image in FIG. 20, the image of the tube 2 marked with a circle 61 can be accurately identified and read even by an inexperienced person, an inexperienced person or an amateur. It becomes possible.

The above first method will be described. In the processing circuit 10, steps a1 to a2 in FIG.
Then, the original image shown in FIG. 4 is read from the memory 11. This original image is shown in simplified form in FIG. The tube 2 is axisymmetric with respect to the vertical line 3,
In the original image of (1), the image 2a of the tube 2 is line-symmetrical with respect to the vertical first symmetry line 13. This original image is stored in the memory 11.

In step a3, difference processing is performed, and in step a4, the secondary image shown in FIG. 22 (2) is formed. In the original image of FIG. 22 (1), reference numeral 14 is soil 1.
The image of the surface of is shown. The image 2a of the tube 2 corresponds to the waveform 15 of one (eg, positive) polarity of the reflected wave shown in FIG. 2 (4), and the white image 16 is obtained, and the other (eg, negative) image is obtained.
Corresponding to the polar waveform 17, a shaded waveform 18 is shown.
Is obtained.

FIG. 23 shows the original image stored in the memory 11. In this original image, in the x, y plane,
The signal level is stored for each pixel in a matrix.
The position Q1 has coordinates (xq1, yq1), its signal level is Lq1, and the position Q2 has coordinates (xq2, y).
q1) and its signal level is Lq2. The signal level L is stored in the store cells at the positions Q1 and Q2.
q1 and Lq2 are stored respectively. Step a
In 3, the following equation 3 is calculated.

[0038]

## EQU00003 ## .DELTA.L (x, y) = L (x, y) -L (x-.DELTA.x, y) where the signal level at the position of coordinates (x, y) is L
(X, y), and the signal level at the position (x-Δx, y) is L (x-Δx, y). The value of the level difference ΔL (x, y) represented by the equation 1 is stored in the store cell at the position (x, y). At the positions Q1 and Q2 described above,

[0039]

## EQU00004 ## xq2 = xq1-.DELTA.x where .DELTA.x is a value at predetermined intervals in the x direction. The x direction is the direction perpendicular to the first line of symmetry, that is, the horizontal direction in this embodiment. Δx is set to a value corresponding to, for example, about 5 cm in the soil 1. Position Q
The difference ΔL between the signal levels Lq1 and Lq2 between one pixel and another pixel at the position Q2 at a predetermined fixed interval Δx in the direction perpendicular to the first symmetry line 13 with respect to the pixel Q1 is The level difference ΔL for each pixel Q1 is obtained based on the equation 1 as shown in the equation 3, and is stored in the store cell of the pixel Q1.

[0040]

## EQU00005 ## .DELTA.L = Lq1-Lq2 In this way, the difference processing is performed and is performed for the pixels of the entire surface of the original image, and the secondary image shown in FIG. 22 (2) is obtained in step a4. This secondary image is stored in the memory 11. In the secondary image in FIG. 22 (2), the positive polarity images 19 and 20 shown in white.
And the negative images 21 and 22 shown by hatching
The symmetry line 13 is formed substantially symmetrically to the left and right, and the polarities of the upper and lower sides are reversed. According to such difference processing, the interference due to the horizontally continuous reception signal is weakened, and the apex 2 of the image 2b of the tube 2 near the first line of symmetry 13 is reduced.
The image of 3 disappears.

Next, in steps a5 to a12 of FIG. 3, the reverse synthetic aperture processing is performed. Therefore, in step a5, as shown in FIG.
The line of the arc 25 shown in is set.

[0042]

(X−x) 2 + Y2 = k · y2 This line 25 is line-symmetric with respect to the second line of symmetry 27 parallel to the first line of symmetry 13, and its vertex position R has coordinates (X, Y). Have. The constant k corresponds to the relative permittivity of the soil 1 in which the pipe 2 is buried. This value k is preset corresponding to the dielectric constant of the soil 1. In other embodiments of the invention, line 25 may be a hyperbola or other arc-shaped function.

Next, as shown in FIG. 22 (4), on the screen 24, the left side line portion 26a and the right side line portion 26b with respect to the second symmetry line 27 of the line 25 are on the respective line portions 26a and 26b. The signal levels of a plurality of pixels in each are added to steps a6 and a7.
To execute. At step a6, as shown in FIG. 24 (1), each coordinate (xpi,
When the signal level L of (ypi) is i = 0 at the vertex R and i = u at the predetermined final end Ra of the line portion 26a, the operation of Expression 7 is performed.

[0044]

[Equation 7]

Similarly, as shown in FIG. 24 (2), the coordinates (xmi, ym on the line portion 26b are displayed.
When the signal level L of i) is i = 0 at the apex R and i = v at the final final end Rb of the line portion 26b, the operation of Equation 8 is performed.

[0046]

[Equation 8]

In this way, the signal levels L of the pixels on the second symmetry line 27 are added at the left and right line portions 26a and 26b of the second symmetry line 27, respectively, and the left sum E1 is obtained.
The sum E2 on the right side is calculated and obtained.

Next, at step a8, these sums E1,
The difference ΔE from E2 is obtained as shown in Equation 9.

[0049]

ΔE = E1−E2 This difference ΔE is stored in the store cell of the vertex R as a value on the pixel of the vertex R of the second symmetry line 27 in step a9.

At step a10, the second line of symmetry 27 is set to x.
8 is displaced in the direction by 1 as shown by the reference numeral 28 in FIG.
The operation of obtaining and storing E is repeated until the amount of shift in the x direction reaches a predetermined value in step a11. Shift the second line of symmetry 27 in the direction opposite to the arrow 28,
Similarly, the operation of obtaining and storing the difference ΔE is similarly repeated.

In this way, the calculation described with reference to FIG. 22 (4) and FIG. 24 is performed on the screen shown in FIG. 22 (2), and then the screen is shifted by 1 in the y direction at step a12. As a result, the tertiary image 29 shown in FIG. 22 (5) is obtained in step a14. Thus, steps a5 to a1
By executing Step 2, the inversion synthetic aperture process is performed, and thus the vertex 23, and thus the R positive open-polarity image 30 and the shaded negative-polarity image 31 are clearly obtained.
In this way, the feature extraction of the vertex R can be effectively performed. Therefore, although the difference processing has no feature extraction effect, it is highly effective in weakening the adverse effects of horizontally continuous reflection images due to the boundary of groundwater and strata, etc. The present invention is implemented by taking advantage of each processing that the feature extraction effect is likely to be deteriorated by the influence of the reverse image to be performed, whereby the apex of the tube 2 can be detected more accurately. The order of moving the x and y positions may be reversed.

The above-mentioned second method is shown in FIG. This FIG. 25 is similar to FIG. 21 showing the first method, but without the difference processing performed in a3 and a4, and instead of step a8 in FIG. 21, the sum E of the sums E1 and E2 is obtained.
Is calculated in step b8 as shown in Expression 10.
Other operations are similar to those of the above-mentioned first method.

[0053]

[Equation 10] E = E1 + E2 The first and second methods may have other configurations.

The invention is not only implemented in connection with underground pipes 2 but also extensively for the detection of objects buried in concrete and for other objects located in hiding places. It can be carried out.

Instead of using the transmitting and receiving antennas 6 and 7 to obtain the original image, the original image may be obtained by another method using ultrasonic waves or the like.

[0056]

As described above, according to the present invention, the pulse signal is transmitted to the oscillating means 4 at the hiding place where the object is provided.
5 and 6, the reflected wave from the object is received and the time difference W1 from the emission to the reception is obtained to obtain the original image of the concealed place including the object, or the original image of the object is obtained by another configuration. , A point portion in a common area in which each point of the binarized image is binarized For each dot-shaped image, the maximum gradation in each point part and the coordinates of the maximum gradation are calculated, and the maximum gradation of the point part in each common area is mutually multiplied or added. The calculation is performed, the calculation results thus obtained are arranged in order from the largest to the smallest, and only the input number of objects to be detected are selected in descending order of the calculation results, and the objects corresponding to the selected common area are selected. Only the image can be identified on the original image, Superimposed display on the basis of the serial coordinates. As a result, even if the person who is unfamiliar, inexperienced person, or layperson who cannot understand whether or not it is the image of the object only by looking at the original image, seeing only the image of the identified object, the position of the object It becomes possible to know exactly.

Further, in order to improve the reliability of the position detection of the object, each point of the binarized point-shaped image is expanded, then the point portion in the common area is obtained, and this point portion is determined. Further expansion, and thus reliability can be improved.

Furthermore, according to the method of obtaining a dot-shaped image, an original image of an axisymmetric object such as an underground pipe is
For example, it is obtained by observing with an underground exploration radar or the like, and then the signal level difference between each pixel and another pixel at a predetermined fixed interval in the direction perpendicular to the first symmetry line is obtained. It is possible to store the level difference for each pixel and perform the difference processing, obtain the secondary image in this way, and suppress the disturbance due to, for example, a horizontally continuous image such as groundwater.

Thereafter, the signal levels of the pixels on a predetermined line which is line-symmetric with respect to the second line of symmetry parallel to the first line of symmetry are added on the left and right sides of the second line of symmetry, and the left and right sums E1 and E2 are added. Is stored as a value on the pixel at the apex of the line, and such an operation is performed on the first line of symmetry with the second line of symmetry parallel to the first line of symmetry. Repeatedly shifting it in the vertical direction, obtaining a dot-shaped image, performing inverted synthetic aperture processing to clearly extract the features of the vertices, so it was installed in a concealed place such as an underground pipe. The detected object can be surely detected.

[Brief description of drawings]

FIG. 1 is a block diagram showing an electrical configuration of an embodiment of the present invention.

FIG. 2 is a waveform chart for explaining the operation of the embodiment of FIG.

FIG. 3 is a flowchart for explaining the operation of the processing circuit 10 according to the embodiment of the present invention.

FIG. 4 is a diagram showing an original image displayed by a display unit 12.

FIG. 5 is a diagram showing stored contents of an original image stored in a memory 11.

FIG. 6 is a diagram showing a first punctiform image obtained by a first method.

FIG. 7 is a diagram showing an image obtained by binarizing a first dot-shaped image.

8 is a diagram showing an image after dilation calculation processing after dilation of the binarized image shown in FIG. 7 into four neighborhoods and eight neighborhoods.

FIG. 9 is a diagram for explaining 4-neighbor expansion.

FIG. 10 is a diagram showing 8-neighbor expansion.

FIG. 11 is a diagram showing a second dot-shaped image obtained by the second method.

FIG. 12 is a diagram showing an image obtained by binarizing a second dot-shaped image.

13 is a diagram showing an image obtained by subjecting the binarized image shown in FIG. 12 to a 4-neighbor expansion and a 8-neighbor expansion to perform a dilation calculation process.

FIG. 14: AN of the dilated image shown in FIGS.
It is a figure which shows the image after D processing.

FIG. 15 is a common area 5 of points 51 and 52 after AND processing.
It is a figure for demonstrating the operation | movement which calculates | requires 3.

FIG. 16 is a diagram showing an image obtained by expanding an AND-processed image.

FIG. 17 is a diagram for explaining expansion calculation processing.

FIG. 18 is a diagram for explaining the necessity of performing expansion calculation processing on the common area 53.

FIG. 19 is a diagram showing the stored contents of the memory 11 for explaining the labeling operation in step 9;

FIG. 20 is a diagram showing a state in which a mark 61, which is an identification symbol, is superimposed on the original image on the display means 12 to facilitate the interpretation corresponding to the tube 2.

FIG. 21 is a first diagram for obtaining a first punctiform image from an original image.
6 is a flowchart for explaining the operation of the processing circuit 10 for obtaining the method.

FIG. 22 is a diagram showing an image for explaining the principle of the first method.

FIG. 23 is a diagram for explaining an operation of difference processing.

FIG. 24 is a diagram for explaining inversion synthetic aperture processing.

FIG. 25 is a second diagram for obtaining a second dot-shaped image from the original image.
6 is a flowchart for explaining the operation of the processing circuit 10 that performs the method.

[Explanation of symbols]

1 soil 2 underground pipes 3 transmitter circuit 6 transmitting antenna 7 Receiving antenna 8 Receiver circuit 10 Processing circuit 11 memory 12 Display means 41-46,56 pixels 47,48 4 Nearly expanded point 51,52 points 53 Common areas 58,59 points 54 common region 53 expanded region

Claims (4)

[Claims] 1. An original image of an object provided at a concealed place is obtained, and a dot-like image having a plurality of gradations is obtained by a plurality of different methods, and each dot-like image is binarized and binarized. The point portion in the common area where the points of the digitized point-shaped image overlap is obtained, and the maximum gradation in each point portion and the coordinates of the maximum gradation are obtained for each point-shaped image, The maximum gradations of the point parts in each common area are mutually calculated, the number of objects to be detected is input, and the calculation results for each common area are sorted in order of magnitude.
It is provided in a concealed place characterized by selecting the input number and displaying only an image of an object corresponding to the selected common area on the original image so as to be superimposed and displayed based on the coordinates. Method for detecting the position of an object. 2. The method according to claim 1, wherein each point of the binarized point-like image is expanded, then the point portion within a common area of the points is obtained, and the point portion is expanded. A method for detecting the position of an object provided in the hiding place described. 3. (a) One of the methods is: (a1) each pixel of the original image, and another predetermined interval in the direction perpendicular to the first symmetry line with respect to the pixel;
The level difference of the signal with respect to two pixels is calculated, the level difference is stored for each pixel, the difference processing is performed to obtain the secondary image, and (a2) based on the secondary image, parallel to the first symmetry line. The signal levels of the pixels on a predetermined line that is line-symmetric with respect to the second symmetry line are added on the left and right sides of the second symmetry line, and the difference ΔE between the sum E1 on the left side and the sum E2 on the right side is calculated.
The difference ΔE is stored as a value on the pixel at the apex of the line, and (a3) the second line of symmetry is parallel to the first line of symmetry
The difference ΔE is obtained by shifting in the direction perpendicular to the line of symmetry, and the operation of storing the difference ΔE as the value on the pixel at the apex is repeated over the entire image to obtain the dot-shaped image. (B) Another method is (b1) A second line parallel to the first line of symmetry based on the original image.
The signal levels of pixels on a predetermined line that is line-symmetric with respect to the line of symmetry are added on the left and right sides of the second line of symmetry,
The sum E of the sum E1 on the left side and the sum E2 on the right side is obtained, and the sum E is stored as a value on the pixel at the apex of the line, and (b2) the second symmetry line is parallel to the first symmetry line. In the first state
2. The concealed place according to claim 1, wherein the dot-shaped image is obtained by repeatedly performing an operation of shifting the symmetry line in the vertical direction to obtain the sum E and storing the sum E as a value on a pixel at the apex over the entire area of the image. Position detection method for an object provided in. 4. Oscillating means 4, 5 for radiating a pulse signal
6, the reflected wave from the oscillating means is received, the time difference W1 from the emission to the reception is obtained to obtain the original image of the concealed place including the object to be detected, and the original image is stored in the memory 11 and displayed. Means 7 to 1 for visually displaying by means 12
2, a punctured image creating means for reading out the original image stored in the memory, obtaining the punctured image by a plurality of different methods, and responsive to an output from the punctured image creating means. And binarizing means for binarizing each dot-shaped image, inputting means 60 for inputting the number of objects to be detected, and binarizing means in response to each output of the binarizing means and inputting means 60. Obtaining a point portion in a common area in which each point of the digitized pointized image overlaps, for each pointed image, obtain the maximum gradation in each point portion and the coordinates of the maximum gradation, Calculation means for mutually calculating the maximum gradation of the point portion in each common area, and selecting the calculation result for each common area in the order of magnitude of the calculation result by the input number, and the output of the calculation means In response to the image of the object corresponding to the selected common area on the original image displayed by the display means 12. In the vicinity, identifying symbols, the position detecting device of an object provided in the hiding place, characterized in that it comprises means for displaying by overlapping the basis of the coordinates.