JP2002005649A - Apparatus and method for processing point group data and recording medium stored with point-group data processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an apparatus and a method where slave measuring-point group data which has a desired data amount and which is thinned out nearly equally is obtained from master measuring-point group data. SOLUTION: Three-dimensional coordinate data on the surface 10a of a object to be measured 10, which is measured by a three-dimensional measuring machine 20 is stored in an internal memory 56 at a processing unit 50 as the master measuring-point group data Dp. A processing part 52 at the processing unit 50 executes readout operation of a point-group data processing program Pm from a hard disk 54. The data Dp is read out from the internal memory 56. Measuring points are thinned out equally by a thinning-out processing operation to be reduced to the measuring points in the desired number, to be stored in the hard disk 54 as the slave measuring-point group data Dc.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a point cloud data processing apparatus and a point cloud data processing apparatus for performing a thinning process on measurement point cloud data corresponding to the surface shape of an object to be measured obtained by, for example, a three-dimensional measuring machine. It is about the method.

[0002]

2. Description of the Related Art Conventionally, the shape of an aspherical surface or a free-form surface of an aspherical lens, a toric lens, or a mold for molding them is measured by a three-dimensional measuring machine.
The three-dimensional measuring machine measures three-dimensional coordinates (or two-dimensional coordinates) at predetermined intervals along the surface of the object to be measured, and outputs coordinate data obtained at each measurement point to a dedicated arithmetic processing device. In the arithmetic processing unit, a series of coordinate data is held in an internal memory or the like as measurement point group data, read out as necessary and subjected to appropriate graphic processing, for example, inspection of parts as described above, processing in CAD, and the like. It is used for generating control data for use, drawing drawings by computer graphics, and the like.

The measuring accuracy of recent three-dimensional measuring machines is 10 ⁻⁹ m.
(= 1 nm), and the number of measurement points becomes enormous accordingly. For example, if a range of about several cm is measured, the number of measurement points reaches tens of thousands. However, the processing capacity of the arithmetic processing unit is relatively low compared to the performance of the three-dimensional measuring machine, and the number of measurement points that can be handled at one time is limited to about several thousand, and the processing speed is significantly reduced. It is necessary to reduce a large amount of data.

[0004] As a countermeasure against such a problem, the measurement range is divided, and a series of processes from measurement to graphic processing are performed, and then the data of the surface shape in each range is synthesized, or the measurement interval is increased. Thus, the number of measurement points can be reduced. However, in the former case, the series of processes from measurement to graphic processing and synthesis require a great deal of time and effort, and since the surface shape data does not always match at the boundary of each measurement range, the reliability of the synthesized data is Is reduced. In the latter case, a shape that accurately matches the original surface shape cannot be extracted from the measurement point group data, and when analysis of a fine shape is required, the measurement must be performed again. Such a technique that requires a plurality of measurement operations is extremely inefficient.

In order to complete the measurement operation only once, coordinate data obtained at the minimum measurement interval is stored and held as parent measurement point group data, and measurement points are thinned out from the parent measurement point group data. It is sufficient to generate the child measurement point group data reduced to a more appropriate data amount.

[0006]

However, if the data is simply thinned out, it is not always possible to reduce the data amount to a desired amount. On the other hand, if the data amount after the thinning is determined in advance, the data cannot be uniformly thinned out. The frequency of the measurement points is biased.

The present invention has been made in view of the above points, and a point cloud data processing apparatus and a processing method capable of reducing measurement point cloud data to a desired data amount and uniformly thinning out measurement points. The purpose is to provide.

[0008]

According to the present invention, there is provided a point cloud data processing apparatus which converts two-dimensional or three-dimensional coordinates of a plurality of measurement points arranged at regular intervals along a surface of an object to be measured into a parent measurement point. Input means for inputting as group data, setting means for arbitrarily setting a target value of the number of measurement points, and substantively thinning out measurement points in steps until the number of measurement points of the parent measurement point group data matches the target value. Means for reducing the number of measurement points that are evenly arranged in the same way, and outputting the two-dimensional or three-dimensional coordinates of the number of measurement points selected by the number of measurement points reducing means that match the target value as child measurement point group data. Means. This makes it possible to easily obtain the data of the measurement points having the desired data amount and being thinned out substantially evenly.

The measurement point number reducing means of the point cloud data processing apparatus can execute the first or second thinning-out processing. Specifically, the first thinning-out processing sets the measurement points before thinning to a target value. The block is divided into the same number of blocks, only one point is extracted from each block, and the rest is thinned out. As a result, the number of measurement points can be reduced at a time to a number that matches the target value. The second thinning processing sets a predetermined number of block constituent points at which the number of measured points that have been thinned does not exceed the difference between the number of measured points before thinning and the target value and is the maximum. Decimate only one point each time. As a result, the number of measurement points can be reduced to a number that matches the target value or a number that is larger and very close to the target value. Either the first or second thinning process is selectively executed based on the number of measurement points before thinning and the target value.

[0010] The measuring point number reducing means of the point cloud data processing device further substitutes the number of measuring points of the parent measuring point cloud data as an initial value of the number of measuring points before thinning out, and further reduces the number of measuring points immediately before the number of measuring points matches the target value. The number of measurement points obtained by the thinning processing may be substituted for the number of measurement points before thinning, and the thinning processing may be repeatedly performed. As a result, even when the number of measurement points cannot be reduced to the target value in one thinning process, it is possible to reliably reduce the number of measurement points to the target value,
The child measurement point group data having the number of measurement points consistent with the target value is always obtained.

In the first decimation process executed in the point cloud data processing apparatus, the measurement points before decimation except for the last one point are equally divided into blocks smaller by one than the target value, and the blocks are separated from the blocks. In each case, the first point is selected and the rest are thinned out. As a result, measurement points that are uniformly arranged while leaving both end points can be selected.

In the second decimation process executed by the point cloud data processing apparatus, a difference between the number of measurement points before decimation and the target value is obtained, and the number of measurement points before decimation is divided by one larger than the difference. The number obtained by adding 1 to the quotient of is determined as the block configuration point, the measurement points before thinning are divided into blocks for each block configuration point from the beginning, and the last one point is decimated from the block excluding the last block to obtain the remaining points. Is selected. Thereby, measurement points thinned out at equal intervals can be selected.

In the decimation process executed by the measurement point number reduction means of the point cloud data processing device, specifically, the remainder obtained by dividing a number smaller by one than the number of measurement points before thinning by a number smaller by one than the target value is obtained. When the remainder is 0, the first thinning process is selected, and when the remainder is 1 or more, the second thinning process is selected.

The measuring point number reducing means of the point cloud data processing device preferably substitutes the number of measuring points of the parent measuring point cloud data as an initial value of the number of measuring points before decimation, and sets the number of points immediately before the number of measuring points matches the target value. The number of measurement points obtained by the thinning processing is substituted for the number of measurement points before thinning, and the thinning processing is repeatedly performed. Thus, even if the survey points could not be reduced to the target value in one thinning process,
It can be reliably reduced to the target value.

Further, in the point cloud data processing method of the present invention, two-dimensional or three-dimensional coordinates of a plurality of measurement points arranged at regular intervals in a predetermined direction along the surface of the object to be measured are input as parent measurement point cloud data. A first step of setting a target value of an arbitrary number of measurement points, and a step of thinning out the measurement points stepwise until the number of measurement points of the parent measurement point group data coincides with the target value. And a fourth step of outputting two-dimensional or three-dimensional coordinates of the measurement points selected by the thinning means as child measurement point group data. And
This makes it possible to easily obtain the data of the measurement points having the desired data amount and being thinned out substantially evenly.

In the recording medium according to the present invention, an input processing routine for obtaining, as parent measurement point group data, two-dimensional or three-dimensional coordinates of a plurality of measurement points arranged at regular intervals in a predetermined direction along the surface of an object to be measured. And a setting processing routine for setting a target value of an arbitrary number of measurement points, and a measurement processing method in which measurement points are thinned out step by step until the number of measurement points of the parent measurement point group data coincides with the target value. A point group data processing program including a thinning processing routine for selecting points and an output processing routine for outputting the two-dimensional or three-dimensional coordinates of the measurement point selected by the thinning means as child measurement point group data is stored. It is characterized by the following. As a result, point cloud data processing can be easily performed even on a general-purpose personal computer, and data of measurement points having a desired data amount and being thinned out almost uniformly can always be obtained.

[0017]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a block diagram schematically showing an entire surface shape analysis system using a point cloud data processing device. This system includes a three-dimensional measuring device 20 that measures three-dimensional coordinates of the surface of the device under test 10 and an arithmetic processing device 50 that performs various data processing based on the three-dimensional coordinates obtained from the three-dimensional measuring device 20. And a monitor device 6 for displaying analysis results and the like
0 and an input device 70 such as a mouse or a keyboard.

The three-dimensional measuring device 20 makes a stylus portion 24 provided at the tip of a probe 22 movable relative to the measured object 10 three-dimensionally in contact with the curved surface 10 a of the measured object 10. The probe 22 is sampled using a laser beam at predetermined intervals while relatively moving the probe 22 in a predetermined direction, in the direction parallel to the X axis in FIG. 1, and the three-dimensional coordinates of each measurement point that is a contact point with the curved surface 10a. P1 (X1, Y1, Z
1), P2 (X2, Y2, Z2),... Pi (Xi,
Yi, Zi) (i is an arbitrary natural number).

The relative movement amount of the probe 22 and the object 10 to be measured, that is, the measurement range, the moving direction and the moving speed of the probe 22, and the like are controlled by the control unit 26. The measurement interval can be freely set by an operator, but it is preferable to measure at a measurement interval of μm, which can reproduce a highly accurate surface shape.

The sampled data, that is, each measurement point P
The three-dimensional coordinate data 1 to Pi are sequentially output from the control unit 26 to the arithmetic processing unit 50 which is a point cloud data processing unit.
When the sampling of the curved surface 10a is completed, the obtained series of three-dimensional coordinate data is converted into a file as parent measurement point group data Dp based on a predetermined format by the arithmetic processing unit 52 and stored in the internal memory 56. The information on the number i of the measurement points is stored in the header area of the file.

Actually, sampling in the X direction is repeatedly performed while the Y position is changed little by little. The curved surface 10a
May be performed not only in the X direction but also in the Y direction.

The hard disk 54 of the arithmetic processing unit 50 has installed therein a point cloud data processing program Pm for performing thinning processing and programs for performing various data processing such as graphic processing and analysis processing. Is read by the arithmetic processing unit 52 and various processes are executed. When the point cloud data processing program Pm is executed, the parent measurement point cloud data Dp is read from the internal memory 56,
The data amount is reduced by the thinning process, converted into child measurement point group data Dc, and stored in the hard disk 54 of the arithmetic processing unit 50. Note that the point cloud data processing program Pm may be stored in a suitable external recording medium such as a CD-ROM, or may be installed in an arbitrary personal computer to perform point cloud data processing as needed.

FIG. 2 is a diagram conceptually showing a screen of the monitor device when the point cloud data processing program Pm is executed. On the left side of the screen, the conversion source, that is, the storage location and file list of the parent measurement point group data Dp to be subjected to the thinning process are shown. A drive selection area 102, a folder selection area 104, a file type selection area 106, and a list display area 108 are arranged in order from the top left of the screen. When the drive, folder, and file type are designated by the input device 70 in each of the three selection areas 102, 104, and 106, only the files that match the designated type are stored in the designated drive and folder in the list display area 108. Is displayed.

On the right side of the screen, the conversion destination, that is, the storage location of the child measurement point group data Dc generated by the thinning process is shown.
2, 114, 116 and a list display area 118 are provided.

At the center of the screen, a conversion button 120 and a data number designation area 122 after thinning are provided. An arbitrary natural number is designated by the input device 70 after thinning out.
When the value is input to 22, the value is set as the target value of the number of measurement points after the thinning process. After specifying a file in the conversion source list display area 108 and highlighting the file (indicated by hatching in the figure), the user presses the conversion button 120 on the input device 7.
If you specify 0 operation, for example, by clicking the mouse,
The parent measurement point group data Dp of the designated file is subjected to the thinning process, and the number of measurement points is reduced to the data number designation area 122 after the thinning.
The data amount is reduced until the target value set in is reached. Then, the file name of the child measurement point group data Dc after the thinning processing is displayed in the list display area 118 of the conversion destination.

Note that a plurality of files can be designated in the list display area 108. In this case, the above-described series of point cloud data processing is sequentially performed on each file.

As described above, in this embodiment, since the operator can specify the target value of the number of measurement points, that is, the number of measurement points of the child measurement point group data Dc, the number of measurement points of the parent measurement point group data Dp differs for each file. However, the data amount can always be reduced to an appropriate value. Therefore, if the target value is set to the maximum value of the number of measurement points that can be processed by the graphic processing program of the arithmetic processing device 50, the processing capability of the arithmetic processing device 50 can always be sufficiently functioned. Also, the operator only needs to specify the file names displayed in the list on the screen,
The operation is simple, and there is no need to be aware of the number of measurement points of the parent measurement point group data Dp, so that the user is not bothered.

Further, since the parent measurement point group data Dp is held as it is in the internal memory 56, a plurality of child measurement point group data Dc having different numbers of measurement points can be generated at any time. Since it can be changed only by re-inputting the numerical value of the designated area 122, one parent measurement point group data D
Generation of a plurality of child measurement point group data Dc from p can be performed very easily. Therefore, the child measurement point group data Dc having the number of measurement points most suitable for the analysis process can be obtained easily and in a short time.

Next, a point cloud data processing method according to the present embodiment will be described. In the point cloud data processing method of the present embodiment, j (j ≦ i) measurement points are extracted by thinning out i measurement points, but the measurement points are extracted at regular intervals according to the numerical relationship between i and j. There are times when you can do it and times when you can't. Therefore, a first thinning-out process for thinning out j measurement points at equal intervals in accordance with the relationship between i and j, and a second thinning-out process for extracting measurement points that are not strictly equal but are substantially evenly thinned out. One of the thinning process.

FIG. 3 is a diagram conceptually showing the first thinning process. FIG. 3A shows an arrangement of measurement points before the first thinning process. Specifically, i _{= 25} measurement points measured at intervals L1 along a predetermined direction (horizontal direction in the figure) Are shown. FIG. 3B is a diagram showing an array of measurement points reduced to j _{= 9} by the first thinning-out process.

In order to thin out evenly from 25 points to 9 points, 25 points are divided into 9 blocks, and 1 block is used for each block.
Only points may be selected, but it is preferable that both ends P1 and P25 be left in order to accurately reproduce the surface shape. Therefore, in the first thinning processing, the 24 measuring points P1~P24 excluding P25, equally divided into j ₌ 8 one ₉ than one less block BL ₁ to BL _8, the last block BL ₉ P25 It consists only of. Then, if decimated remaining selects only the beginning of the measurement points in each block BL ₁ to BL _9, the number of measurement points becomes nine. Block B
The number of measurement points constituting each of L _{1 to} BL ₈ , that is, the number of block constituent points Sa is calculated by equation (1), and in this specific example, Sa = 3. Sa = (i-1) / (j-1) (1)

Therefore, in the blocks BL1 to BL8, the remaining two points other than the head are thinned out, so that the thinned data is P1, P4, P7, P10, P13, P1.
6, 9 combining P19 and P22 with the last P25
And the distance between two adjacent measurement points is L1 ×
It becomes 3.

As described above, in the first thinning-out processing, j measurement points arranged at equal intervals in one direction can be selected without losing the measurement points at both ends from the i measurement points. .

FIG. 4 is a diagram conceptually showing the second thinning process. FIG. 4A shows an arrangement of measurement points before the second thinning-out process. Specifically, i _{= 15} measurement points measured at regular intervals along a predetermined direction (horizontal direction in the figure) Are shown. FIG. 4B is a diagram illustrating an array of measurement points reduced to j _{= 12} by the second thinning-out process. FIG. 4C is a comparative example, and is a diagram showing an array of measurement points reduced to j _{= 12} by another process.

In order to thin out from 15 points to 12 points,
If the first 14 points can be equally divided into 11 blocks, the first thinning process can be adopted. However, since the quotient obtained by dividing 14 by 11 is not an integer, it is possible to thin out evenly while leaving both ends as shown in FIG. Can not. Therefore, in the second thinning-out process, the difference Sb between the number i of measurement points and the target value j, that is, the number of measurement points to be deleted is noted, and the i measurement points are determined by the difference Sb.
The number of measurement points per block that can be divided into blocks of the number obtained by adding 1 to the number of blocks, that is, the number of block constituent points Sc is determined, and blocks are sequentially divided from the top, and only the last point is deleted from each block. We take the technique of thinning out only points. Note that if the number of measurement points of the last block is less than the number of block constituent points Sc, it is not deleted from the last block.

Here, as shown in FIG. 4C, if the last point of each block is simply thinned out by dividing the number of blocks corresponding to the difference Sb (= 3) into P5, P10, and the final point P15 Is deleted, the frequency of the measurement points is biased toward the leading side (left side in the figure), and is not evenly thinned out. However, according to the second decimation process, the number of blocks is set to 4, which is one greater than the difference Sb, the measurement points P1 to P15 are divided into four blocks every four points from the top, and three intermediate points P4, P8, and P12
Is decimated. As is clear from comparison with the comparative example shown in FIG. 4C, the measurement points after the thinning processing are arranged uniformly.

As described above, when the number of blocks is set to a value obtained by adding 1 to the difference Sb and when the last measurement point of each block is thinned out, the number of measurement points of the final block BK _{Sb + 1} is less than the number of block constituent points Sc. There is a high probability that measurement points will not be deleted. Therefore, in the second thinning-out process, it is possible to thin out evenly by the number of measurement points that match the difference Sb.

The number Sc of block constituent points is obtained by the following equations (2) and (3). [X] indicates the integer part of the real number X. In equation (3), the maximum value of the number of measurement points that can generate (Sb + 1) blocks BK is obtained, and the number obtained by adding 1 to the integer part of this maximum value, that is, the first decimal point of the maximum value
The number rounded up is determined as the number of block constituent points Sc. Sb = ij (2) Sc = [i / (Sb + 1)] + 1 (3)

In the example shown in FIG. 4, the number of measurement points i is 15,
The standard value j is 12, the difference Sb = 3 according to the equation (2),
The number of block constituent points Sc = 4 is obtained from the equation (3).
You. Therefore, 15 points are the blocks BK four points at a time from the beginning.₁
~ BK_FourDivided into blocks BK ₁~ BK_ThreeFrom the end
Is deleted, and the final block B composed of three points
K_FourOnly not deleted. Adjacent as in the first thinning process
The distance between two matching measuring points is not the same, but the bias of the measuring points
Is prevented.

However, when the number of block constituent points Sc is determined by the equation (3), the following problems occur. For example, when thinning out 50 points to 30 points, the number of block constituent points Sc is three. If you divide it into blocks BK every three points,
Since the number of blocks is 17, and the final block BK is composed of only two points, if the third measurement point of each block BK is decimated, the decimated number is 16 points, and the difference Sb _{= 20} is not enough.
Four points remain and the desired number of points cannot be reduced to 30. As described above, when the difference Sb is a value that is relatively larger than the number of block constituent points Sc, the number of block constituent points Sc is obtained by rounding up, so that the number of blocks to be thinned out, that is, the number of measurement points i is represented by the number of block constituent points Sc. The quotient [i / Sc] at the time of division is the desired number of blocks (Sb + 1).
A smaller scene occurs.

Therefore, if the number of measurement points i is not reduced to the target value j, the first or second thinning processing is performed again to reduce the number of measurement points until the number of measurement points reaches the target value j.
In the case of thinning out 50 points to 30 points as in the above specific example, first, every three points are thinned out to reduce from 50 points to 34 points and further from 34 points to 30 points. Performing the thinning process stepwise in this way is called multi-order processing.
The first thinning process is called a primary process, and the next thinning process is called a second process.

FIG. 5 is a diagram conceptually showing multi-order processing. FIG. 5A shows an arrangement of measurement points before the thinning processing. Specifically, i _{= 50} measurement points P1 measured at regular intervals along a predetermined direction (horizontal direction in the figure). ＰP50 are shown. FIG. 5B is a diagram showing an array of measurement points reduced to 34 by the primary processing, and FIG.
FIG. 8 is a diagram showing an array of measurement points reduced to the target value j _{= 30} by the secondary processing.

First, the number of block constituent points Sc is determined to be 3 by the equations (2) and (3). Then, 50 measurement points P1 to P50 are divided into 17 blocks B every three points from the top.
It is divided into K ₁ ~BK _17. The last block BK ₁₇ is composed of two points (FIG. 5A). Then, to remove the last of the measuring points from the first 16 block BK ₁ ~BK _16, that is thinned out only one point for each third point. Thus, the number of measurement points is reduced to a value 34 closest to the target value 30 and larger than the target value 30 (FIG. 5B).

Here, the second thinning process is again employed to thin 34 points to 30 points. The reason for this is that the quotient obtained by dividing 33 by 29 is not an integer, and the first thinning process cannot be applied. First, i = 34 and j = 30 are substituted into Expressions (2) and (3), and the number of block constituent points S
c = 7 is obtained. And 34 measurement points are 7 from the top
For each point is divided into five blocks BK _'1 ~BK' _5, the last block BK _'5 only consists of six points. If one point is thinned out every seventh point from the beginning, the first four blocks BK ' ₁
.. BK ′ ₄ , the last point is deleted, but not from the last block BK ′ ₅ . Therefore, the number of measurement points is made to coincide with the target value 30 (FIG. 5C). Since the number of measurement points has become equal to the value of the target value j in the secondary processing, no further thinning processing is performed.

As described above, in the case of the second thinning processing, the desired number of measurement points may not be obtained only by the first processing, but in this case, the second, third,... Is performed step by step, the measurement points can be thinned out almost always and to a desired number. In the present embodiment, the desired number of measurement points is obtained in the secondary processing, but the processing may continue in the third and fourth orders.

Next, the main routine of the point cloud data processing program Pm will be described with reference to the flowchart of FIG.

First, when the point cloud data processing program Pm is started, the screen shown in FIG. 2 is displayed on the screen of the monitor device 60 in step S102. When the file name of the parent measurement point group data Dp to be subjected to the thinning process is specified in step S104, the specified file is read,
The recorded data, that is, the three-dimensional coordinate data of the measurement points, the number of measurement points, and the like are temporarily stored in the RAM of the arithmetic processing unit 52.

Next, step S106 is executed, and the number of measurement points of the specified file is given as an initial value to a parameter i indicating the number of measurement points before the thinning processing. In step S108, the target value of the number of measurement points to be obtained after the thinning processing is input to the post-thinning data number designation area 122 by the input device 70, and this target value is substituted for the parameter j.

When initial values are given to parameters i and j in steps S106 and S108, step S
Go to 110, where the value of parameter i is
It is determined whether the value is greater than the value of. If the value of the parameter i is smaller or equal, it is considered that the current number of measurement points is equal to or smaller than the target value and it is not necessary to perform the thinning process, and the process proceeds to the file saving process of step S114.

When it is determined in step S110 that the value of the parameter i is larger, that is, the current number of measurement points is larger than the target value, a thinning subroutine of step S200 is executed to reduce the number of measurement points to the target value or less. You. When the decimation process is completed, the measured points after the decimation are substituted for the parameter i in step S112, and the process returns to step S110 to determine the magnitude of the updated parameter i and the target value j. In short, when the number of measurement points is larger than the target value, the thinning process is repeatedly performed until the number of measurement points becomes equal to or less than the target value.
When the value becomes equal to or smaller than the target value, the process proceeds to the file saving process in step S114.

In step S114, three-dimensional coordinate data corresponding to the selected measurement point, that is, a measurement point that has not been thinned out, is extracted from the coordinate data stored in the RAM, and is set as child measurement point group data Dc. Save as one file based on a predetermined format. The format of the child measurement point cloud data Dc is the parent measurement point cloud data Dp
The format may be the same as in the case of. In addition, a file name may be automatically generated when saving, or may be arbitrarily input by an operator.

FIG. 7 is a flowchart showing details of the thinning-out processing subroutine (step S200) shown in FIG.

In step S202, 1 is obtained from the number of measurement points i.
The quotient Sa and the remainder R when the number smaller than the target value is divided by the number smaller than the target value by 1 are obtained. Next step S20
In 4, it is determined whether the remainder R is 0, that is, whether the (i-1) measurement points excluding the last measurement point can be equally divided for each Sa. When the remainder R is 0 as in the specific example shown in FIG. 3, it is determined that the division can be performed equally, excluding both ends, the quotient Sa is set to the number of block constituent points, and the first thinning process in step S206 is performed. Be executed.

In step S206, the (i-1) measurement points obtained by removing the last measurement point from the i measurement points are divided into blocks BL for each Sa, and the first of each of the (j-1) blocks BL is divided. Is selected, and all remaining measurement points are deleted. That is, the (Sa × k + 1) th (k: 0 ≦ k ≦ j−
Only (1, integer) measurement points are selected, and as a result, an array of measurement points having the same number of measurement points as the target value j and arranged at equal intervals is obtained.

On the other hand, when the remainder R is an integer greater than or equal to 1 and less than (j-1) as in the specific examples shown in FIG. 4 or FIG. 5, the remainder R is not 0 in step S204.
In other words, it is determined that it is not possible to select measurement points arranged at equal intervals while leaving both ends, and steps S208 to S212 are performed.
(See the specific example shown in FIGS. 4 and 5).

In step S208, the difference Sb between the parameter i and the parameter j is obtained by the above equation (2), and in step S210, the number of block constituent points Sc is obtained by the above equation (3). In step S212, the measurement points are divided into blocks BK every Sc from the beginning, only the last point of each block BK excluding the last block is thinned out, and the rest is selected, that is, Sc × m-th (m : 1 ≦ m ≦ Sb) Only the measurement points are thinned out, and as a result, the thinning interval becomes substantially constant in the arrangement of the measurement points.

When the first or second thinning-out processing is completed, this subroutine is completed, and step S shown in FIG.
Return to the main routine to execute 112.

As described above, in the point cloud data processing apparatus according to the present embodiment, the parent measurement point cloud data Dp is read, and the first or second thinning-out processing is selectively and repeatedly executed to obtain the number of measurement points i. Can be reduced stepwise until the desired target value j is reached, and can be output as child measurement point group data Dc.

[0060]

As described above, according to the present invention, the point data of the measurement point group data can be evenly thinned out, and the processing capability of the arithmetic processing unit can always be maximized.

[Brief description of the drawings]

FIG. 1 is a block diagram schematically showing a surface shape analysis system using a point cloud data processing apparatus according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram showing a monitor screen when a point cloud data processing program is executed.

FIG. 3 is a diagram conceptually showing a point cloud data processing method of the present invention, and is a specific example showing a first thinning process.

FIG. 4 is a diagram conceptually showing a point cloud data processing method of the present invention, and is a specific example showing a second thinning process.

FIG. 5 is a diagram conceptually showing a point cloud data processing method of the present invention, which is a specific example showing multi-order processing.

FIG. 6 is a flowchart showing a main routine of a point cloud data processing program executed in the point cloud data processing device.

FIG. 7 is a flowchart showing details of a thinning-out processing subroutine shown in FIG. 6;

[Explanation of symbols]

 Reference Signs List 10 DUT 50 Arithmetic processing unit (point group data processing unit) 52 Arithmetic processing unit (measurement point number reduction means) P1 to Pi Measurement points Dp Parent measurement point group data Dc Child measurement point group data Pm Point group data processing program

Claims (8)

[Claims] 1. An input means for inputting two-dimensional or three-dimensional coordinates of a plurality of measurement points arranged at regular intervals in a predetermined direction along a surface of an object to be measured as parent measurement point group data, and a target value of the number of measurement points Setting means for arbitrarily setting the measurement points; selecting the measurement points arranged substantially evenly by thinning out the measurement points stepwise until the number of measurement points of the parent measurement point group data matches the target value It is characterized by comprising: measuring point number reducing means; and output means for outputting, as child measuring point group data, two-dimensional or three-dimensional coordinates of the measuring points selected by the number matching the target value by the measuring point number reducing means. Point cloud data processing device. 2. A first decimation process in which the measurement point number reduction means divides the measurement points before decimation into blocks each having the same target value, extracts only one point from each block, and decimates the rest. The predetermined number of block constituent points is set so that the number of measured points before thinning does not exceed the difference between the number of measured points before thinning and the target value and is maximum, and the number of measurement points before thinning is one point from the beginning for each block constituent number. A second thinning-out process for thinning out only one of the first and second thinning-out processes is selectively executed based on the pre-thinning-out measurement points and the target value. The point cloud data processing device according to claim 1, wherein 3. The measurement point number reduction unit substitutes the measurement point number of the parent measurement point group data as an initial value of the measurement point number before the thinning-out, and performs a thinning process immediately before the measurement point number matches the target value. 3. The point group data processing apparatus according to claim 2, wherein the number of measurement points obtained by the above is substituted for the number of measurement points before thinning, and the thinning processing is repeatedly performed. 4. The first decimation process divides the measurement points before decimation excluding the last one point into equal number of blocks smaller by one than the target value, and each of the measurement points is separated from the block by a leading one. 3. The point cloud data processing apparatus according to claim 2, wherein the points are selected and the rest are thinned out. 5. The second decimation process obtains a difference between the number of measurement points before decimation and the target value, and divides the number of measurement points before decimation by one greater than the difference to obtain a quotient of one.
Is determined as the number of block constituent points, and the measurement points before the thinning are divided into blocks for each of the block constituent points from the beginning, and the last one point is thinned out from each block excluding the last block to select the rest. 3. The point cloud data processing apparatus according to claim 2, wherein: 6. In the thinning-out process by the measurement point number reduction means, a remainder is obtained when a number smaller than the number of measurement points before thinning by one is divided by a number smaller than the target value by one, and the remainder is zero. 3. The point cloud data processing apparatus according to claim 2, wherein the first thinning process is selected, and when the remainder is 1 or more, the second thinning process is selected. 7. A first step of inputting, as parent measurement point group data, two-dimensional or three-dimensional coordinates of a plurality of measurement points arranged at regular intervals in a predetermined direction along the surface of the object to be measured; A second step of setting a target value of the points; and the measurement points arranged substantially uniformly by thinning out the measurement points stepwise until the measurement points of the parent measurement point group data match the target values. A point group comprising: a third step of selecting a point; and a fourth step of outputting two-dimensional or three-dimensional coordinates of the measurement point selected by the thinning means as child measurement point group data. Data processing method. 8. An input processing routine for obtaining, as parent measurement point group data, two-dimensional or three-dimensional coordinates of a plurality of measurement points arranged at regular intervals in a predetermined direction along the surface of the measured object; A setting processing routine for setting a target value; selecting the measurement points arranged substantially evenly by thinning out the measurement points stepwise until the number of measurement points of the parent measurement point group data matches the target value Storing a point group data processing program comprising: a thinning processing routine for performing a thinning process; and an output processing routine for outputting two-dimensional or three-dimensional coordinates of the measurement point selected by the thinning means as child measurement point group data. Recording medium.