JP2016038669A - Projection image generation device, projection image generation program and projection image generation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a projection image generation device that easily obtains two-dimensional coordinate data on a projection image having a broad-range and highly accurate three-dimensional shape of an object projected.SOLUTION: A projection image generation device generating two-dimensional coordinate data on a projection image having a surface of an object projected on a two-dimensional plane surface from three-dimensional coordinate data on a distance image representing a three-dimensional shape of the surface thereof is configured to have: an acquisition unit that acquires the three-dimensional coordinate data; a cross-section setting unit that sets a plurality of cross sections apart from each other at a prescribed pitch in a specific direction of the object and parallel to each other therein; an approximation curve identification unit that identifies an approximation curve approximate to the shape of the surface included in the cross section, using the three-dimensional coordinate data; a two-dimensional coordinate data calculation unit that calculates two-dimensional coordinate data on a point on the basis of a distance from a reference point on the approximation curve for each point on the surface included in the cross section; and a storage unit that stores the two-dimensional coordinate data.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a projection image generation apparatus, a projection image generation program, and a projection image that generate two-dimensional coordinate data of a projection image obtained by projecting the surface of an object onto a two-dimensional plane from three-dimensional coordinate data representing the three-dimensional shape of the object. It relates to the generation method.

  In various industries, measuring a three-dimensional shape of an object is performed. There are various three-dimensional measuring devices for measuring a three-dimensional shape, and the characteristics of each device differ depending on the purpose of measurement, but many devices are on the surface of an object that is the object of measurement on the principle of triangulation. Measure the distance to each point and collect point cloud measurement data. Typical measurement methods include a light cutting method using laser light and a correlation method using a stereo camera.

  These three-dimensional measurement devices obtain measurement data of measurement point groups divided in a grid in a field of view that radiates from the projection center. The width of the field of view and the density of measurement points depend on the performance of the three-dimensional measurement device. To do.

  If the measurement target surface is within the field of view of the three-dimensional measurement apparatus, and the density of the measurement points is sufficient and the measurement density is high, the measurement points necessary for measuring the three-dimensional shape can be obtained by one measurement. The measurement points measured in this way have three-dimensional coordinates or a two-dimensional lattice index along with the distance. Such an index is equivalent to two-dimensional coordinate data. By having two-dimensional coordinate data, when calculating a local geometric characteristic, the convenience that a point in the vicinity of a measurement point can be easily searched by referring to the two-dimensional coordinate is generated. .

  However, when the above conditions are not satisfied, for example, when the measurement target surface does not fit in the field of view of the three-dimensional measurement apparatus, or when measurement cannot be performed from one place due to shielding, the entire object can be measured by one measurement. It becomes impossible to measure a three-dimensional shape. Even if measurement is not impossible, the measurement condition may deteriorate and the accuracy of the measurement result may decrease. In such a case, instead of a single measurement, a single three-dimensional measuring device is moved to a plurality of viewpoints to measure from each viewpoint, or a plurality of three-dimensional measuring devices are used from different viewpoints. By measuring, the entire three-dimensional shape of the object can be measured.

  Thus, when there are a plurality of independent three-dimensional coordinate data, unlike the single three-dimensional coordinate data, the whole cannot be expressed by one two-dimensional coordinate, and the convenience of the proximity search using the two-dimensional coordinates is lost. Is called. Therefore, the same convenience as that of single three-dimensional coordinate data can be obtained by reconstructing the two-dimensional coordinate data for a plurality of three-dimensional coordinate data.

  When converting 3D coordinate data into 2D coordinate data, for example, a cross section of an object to be converted is set, and the square of the distance from each of a plurality of points on the cross section is set using the least square method. An approximate curve calculated so as to minimize the sum is used (for example, see Patent Document 1).

  Here, in order to improve the accuracy of approximation when approximating a three-dimensional shape with an approximate curve, set the cross section in the direction according to the shape of the object (set the cross section direction for each cross section), or It is conceivable to calculate a large number of approximate curves by increasing the number of cross sections to be set.

  However, if the direction of the cross section is set for each cross section according to the shape of the object, the two-dimensional coordinate data calculated for each cross section must be combined in consideration of the direction of each cross section. It becomes complicated. Further, when the number of cross sections set for an object increases, the amount of information becomes enormous and information processing becomes complicated.

Japanese Patent No. 4168758

  The present invention has been made in order to solve the above-described problems of the prior art, and can easily obtain two-dimensional coordinate data of a projected image obtained by projecting a wide-range, high-density three-dimensional shape of an object. It is an object of the present invention to provide a projected image generating apparatus, a projected image generating program, and a projected image generating method.

  The present invention is a projection image generation device that generates two-dimensional coordinate data of a projection image obtained by projecting a surface onto a two-dimensional plane from three-dimensional coordinate data of a distance image representing a three-dimensional shape of the surface of an object, An acquisition unit that acquires three-dimensional coordinate data, a cross-section setting unit that sets a plurality of cross-sections parallel to each other at a predetermined pitch in a specific direction of the object, and an approximate curve that approximates the surface included in the cross-section An approximate curve specifying unit that is specified based on the three-dimensional coordinate data of the point, and for calculating the two-dimensional coordinate data of the point based on the distance from the reference point on the approximate curve for each point on the surface included in the cross section 2 A dimensional coordinate data calculation unit and a storage unit for storing two-dimensional coordinate data are provided.

  According to the present invention, it is possible to easily obtain two-dimensional coordinate data of a projected image obtained by projecting a wide-range, high-density three-dimensional shape of an object.

It is a block diagram which shows embodiment of the projection image generation apparatus concerning this invention. It is a flowchart which shows embodiment of the projection image generation method concerning this invention. It is a schematic diagram which shows a mode that the distance image of an object is acquired. It is a schematic diagram which shows the example of the normal vector of each point on the surface of an object. It is a schematic diagram which shows a mode that the said normal vector was distributed on the unit sphere. It is a schematic diagram showing a normal line of a plane passing through the origin with respect to the end point point group of the normal vector. FIG. 6 is a schematic diagram illustrating an example of an initial basic axis calculated based on a three-dimensional coordinate distribution of an object and two basic axis candidates calculated by deviating from the initial basic axis by a minute angle. It is a schematic diagram which shows the example of the cross section set based on said one basic axis candidate. It is a schematic diagram which shows the example of the cross section set based on the said other basic axis candidate. It is a schematic diagram which shows the position of several cross-sections set mutually parallel to an object. It is a schematic diagram which shows the reference direction in which the cross section of an object is set. It is a schematic diagram which shows the example of the surface of the object which appears in the cross section set to the object. It is a schematic diagram which shows the example of the approximated curve which approximates the shape of the surface of the object which appears in the said cross section. It is a schematic diagram which shows the example of another approximated curve which approximates the shape of the surface of the object which appears in the said cross section. It is a schematic diagram which shows the center and radius of a circle as an approximated curve. It is a schematic diagram showing an angle formed by a line connecting the center of the circle and a reference point on the circle and a line connecting the center of the circle and a point on the surface of the object. It is a schematic diagram which shows the example of the range (sampling window) of the surface of the said object by which the said approximated curve is specified. It is a schematic diagram which shows the example of the approximated curve specified for every adjacent cross section. It is a schematic diagram which shows another example of the approximated curve specified for every adjacent cross section.

  Embodiments of a projected image generation apparatus, a projected image generation program, and a projected image generation method according to the present invention will be described below with reference to the drawings.

  In the following, an example in which two-dimensional coordinate data of a projected image obtained by projecting the surface of the object M onto a two-dimensional plane is calculated from the three-dimensional coordinate data of the surface of the object M will be described.

Projection image generator
FIG. 1 is a block diagram showing an embodiment of a projected image generating apparatus according to the present invention. In the projected image generation apparatus according to the present invention (hereinafter referred to as “the present apparatus”), the projected image generation program according to the present invention (hereinafter referred to as “the present program”) cooperates with the hardware constituting the apparatus, A projected image generation method (hereinafter referred to as “the present method”) according to the present invention is executed. This apparatus is realized by, for example, a personal computer.

  In addition, by operating the program on a computer different from the apparatus, it is possible to cause the computer to function in the same manner as the apparatus and to execute the method.

  The apparatus 1 includes a three-dimensional coordinate data acquisition unit 11, a reference direction determination unit 12, a cross-section setting unit 13, an approximate curve specifying unit 14, a two-dimensional coordinate data calculation unit 15, and a storage unit 16.

  The three-dimensional coordinate data acquisition unit 11 is means for the apparatus 1 to acquire three-dimensional coordinate data of the object M. A method for acquiring the three-dimensional coordinate data will be described later.

  The reference direction determination unit 12 is means for determining a reference direction. The reference direction is a direction used when a cross section is set on the object M. A method for setting the reference direction will be described later.

  The cross-section setting unit 13 is means for setting a cross-section for the object M. A method for setting the cross section will be described later.

  The approximate curve specifying unit 14 is a means for specifying an approximate curve that approximates the shape of the surface of the object M included in the set cross section. A method for specifying the approximate curve will be described later.

  The two-dimensional coordinate data calculation unit 15 is means for calculating two-dimensional coordinate data of a point on the surface of the object M using the identified approximate curve. A method for calculating the two-dimensional coordinate data will be described later.

  The storage unit 16 is means for storing information necessary for the apparatus 1 to execute the method. The information stored in the storage unit 16 includes a three-dimensional coordinate database DB1 and a two-dimensional coordinate database DB2. DB1 stores the three-dimensional coordinate data of the object M acquired by the apparatus 1. DB2 stores the two-dimensional coordinate data of the object M calculated by the apparatus 1.

● Projection image generation method ●

Next, an embodiment of this method will be described.
FIG. 2 is a flowchart showing an embodiment of the present method.

First, the present apparatus 1 acquires three-dimensional coordinate data of an object (S1).
FIG. 3 is a schematic diagram showing an example of acquisition of three-dimensional coordinate data of the object M, and images the surface (appearance) of the object M with three imaging devices Cam1, Cam2, and Cam3 arranged at different positions. The example which acquires the distance image of the object M is shown. Reference numerals A1, A2, and A3 indicate imaging regions on the surface of the object M that are imaged by the imaging devices Cam1, Cam2, and Cam3, respectively.

  A distance image is an image in which depth information (information about the distance from the imaging device of the imaging device to the subject) is stored in each pixel constituting the image. The imaging device Cam (Cam1, Cam2, Cam3) is a time-of-flight method that acquires a distance image from, for example, a round-trip time of a laser between the object M and the imaging device Cam when the object M is irradiated with a laser or the like. It obtains by the well-known method. The apparatus 1 can specify the three-dimensional coordinate data of the object M from the distance image. Therefore, the present apparatus 1 can obtain three-dimensional coordinate data of the entire object M by synthesizing a plurality of distance images of the object M obtained by imaging from a plurality of viewpoints (a plurality of imaging devices).

  In the figure, the object M is imaged by three imaging devices, but the number of imaging devices used to acquire the distance image of the object M is not limited to three.

  The apparatus 1 stores the acquired three-dimensional coordinate data of the object M in the DB 1.

Returning to FIG.
Next, the apparatus 1 determines the reference direction using the reference direction determination unit 12 (S2). There are various methods for determining the reference direction. For example, (1) a method using the distribution of normals on the surface of the object M (hereinafter referred to as “determination method 1”), (2) described below. There are two methods: a method using an evaluation function (hereinafter referred to as “determination method 2”). However, in the present invention, the method for determining the reference direction is not limited to these two methods.

(1) Reference direction determination method 1
A reference direction determination method 1 executed by the apparatus 1 will be described.
First, the device 1 specifies the direction of the normal vector of each point based on the three-dimensional data of each point on the surface of the object M stored in the storage unit 16.
FIG. 4 is a schematic diagram showing the direction of the normal vector of each point on the surface of the object M.

Next, the apparatus 1 distributes the normal vectors of the points on the identified object M on the unit sphere. Distributing the normal vector on the unit sphere means that the normal vectors are arranged so that the start point of each normal vector is located at the center (origin) of the unit sphere.
FIG. 5 is a schematic diagram showing normal vectors of points on the object M distributed on the unit sphere.

Next, the present apparatus 1 applies a plane passing through the origin of the unit sphere to the end point group of each normal vector, and determines the normal of the plane as a reference direction.
FIG. 6 shows a plane (indicated by a dotted line) passing through the origin of the unit sphere applied to the end point group of each normal vector, and a normal line of the plane (indicated by an arrow in the figure). FIG.

(2) Reference direction determination method 2
Next, a reference direction determination method 2 executed by the apparatus 1 will be described.
First, the apparatus 1 calculates (specifies) the axial direction of the initial basic axis based on the distribution of the three-dimensional data of each point on the surface of the object M stored in the storage unit 16. Here, for example, the above-described reference direction determination method 1 may be used to calculate the axial direction of the initial basic axis. That is, the normal vector of each point on the surface of the object M is distributed on the unit sphere, a plane passing through the origin of the unit sphere is applied to the point group of the end point of each normal vector, and the normal of the plane is set. The axial direction of the initial basic axis. After that, the present apparatus 1 calculates the axial direction of the basic axis candidate in which the axial direction of the initial basic axis is displaced by a minute angle. The minute angle and the number of basic axis candidates are stored in the storage unit 16 in advance.
FIG. 7 is a schematic diagram illustrating the axial direction SP of the initial basic axis, the axial direction SA of the first basic axis candidate, and the axial direction SB of the second basic axis candidate calculated by the apparatus 1. FIG.

Next, the apparatus 1 sets a plurality of cross sections parallel to the object M for each basic axis candidate. A method for setting the cross section will be described later.
FIG. 8 is a schematic diagram showing the position of the cross section of the object M set in the axial direction of the basic axis candidate SA. On the other hand, FIG. 9 is a schematic diagram showing the position of the cross section of the object M set in the axial direction of the basic axis candidate SB.

  Next, the apparatus 1 specifies an approximate curve of the shape of the surface of the object M included in the cross section for each set cross section using the three-dimensional coordinate data stored in the DB 1 for each basic axis candidate. Thus, an approximate curve group is obtained. A method for specifying the approximate curve will be described later.

  Next, the apparatus 1 evaluates the approximate curve group specified for each basic axis candidate with the evaluation function, and determines the axial direction of the basic axis candidate that has obtained the highest evaluation function value as the reference direction.

  Here, as an example of the evaluation criteria of the evaluation function, for example, the distance between the point group and the approximate curve, that is, “a point on the surface of the object M included in the cross section in which the approximate curve is set” and “at this point” The distance between the corresponding points on the approximate curve and the parameter difference between the approximate curves (for example, the difference in the radius of the circle as the approximate curve).

  The apparatus 1 determines the reference direction used when setting the cross section for the object M by the determination method 1 described above, the second method, or the other method.

  In the following description, for convenience of description, the reference direction is the X-axis direction as shown in FIG.

Returning to FIG.
Next, the apparatus 1 sets a plurality of cross sections parallel to each other at a predetermined pitch in the previously determined reference direction (S3). That is, the shape of the surface of the object M included in the cross section set by the apparatus 1 appears on the YZ plane.

  FIG. 10 is a schematic diagram showing the position of the cross section set by the apparatus 1. In the figure, six cross sections CS1, CS2,..., CS6 are set in the X-axis direction for the object M, and the X coordinates of each cross section are X1, X2,. In other words, each cross section is parallel to each other. The figure also shows that the interval (pitch) between the cross sections is P. The interval P between the cross sections is stored in the storage unit 16 in advance.

  Next, the apparatus 1 specifies an approximate curve that approximates the shape of the surface of the object M included in the cross section (the shape on the YZ plane) for each set cross section (S4).

  FIG. 12 is a schematic diagram showing the relationship between the cross section set for the object M and the shape of the surface of the object M included in the cross section. The figure shows the shape of the surface of the object M included in the cross section set to the X coordinate Xi and the shape of the surface of the object M included in the cross section set to the X coordinate Xj.

  Here, the apparatus 1 specifies an approximate curve that approximates the surface of the object M included in the cross section, using the three-dimensional coordinate data stored in the DB 1. As a method for the apparatus 1 to specify the approximate curve, a least square method or other known methods are used. That is, for example, when the least square method is used, a curve calculated so that the sum of the squares of the distances (deviations) from each point on the surface of the object M included in the cross section is specified as an approximate curve. .

  FIG. 13 is a schematic diagram showing an example of an approximate curve that approximates the surface of the object M included in the cross section set to the X coordinate Xi specified by the apparatus 1. This figure shows that a circle (circular arc) having a center Ci and a radius ri is specified as an approximate curve.

FIG. 14 is a schematic diagram showing an example of an approximate curve that approximates the surface of the object M included in the cross section set to the X coordinate Xj specified by the apparatus 1. This figure shows that a circle (its arc) having a center Cj and a radius rj is specified as an approximate curve as an approximate curve.

  Thus, this apparatus 1 specifies an approximate curve for every cross section. Therefore, for example, when the shape of the approximate curve is a circle, the line connecting the centers of the circles as the approximate curve specified for each cross section is a non-straight line, and the length of the radius of each circle is also different.

Returning to FIG.
Next, the present apparatus 1 is configured for each point on the surface of the object M based on the distance from the reference point on the approximate curve set for the cross section for each point on the surface of the object M included in the cross section. Is calculated (S5).

FIG. 15 is a schematic diagram showing an example of a circle as an approximate curve identified by the apparatus 1 at the X coordinate Xi, and shows that the center of the circle is Ci and the radius of the circle is ri.
FIG. 16 shows a line t0 connecting the center Ci of the circle and a reference point (not shown) on the circle and a point P on the surface of the object M in the circle as the approximate curve shown in FIG. It is a schematic diagram which shows the angle which the line t which connects and forms.

  The apparatus 1 uses the two-dimensional coordinate data (u, v) as the u-coordinate value as the X-coordinate value in which the cross section is set, as defined by the following formula 1, and on the surface of the object M included in the cross-section. The geodetic distance on the approximate curve when the point group is projected onto the approximate curve is calculated as the v coordinate value. That is, when the approximate curve is an arc of a circle, the present apparatus 1 converts the two-dimensional coordinate data for each point on the surface included in the cross section into a circle radius and a “line connecting the center of the circle and the reference point”. And the angle formed by the “line connecting the center of the circle and the point”,

● Summary According to the embodiment described above, in order to calculate the two-dimensional coordinate data by specifying the approximate curve for each cross section, the two-dimensional coordinates of the projected image obtained by projecting the wide-range, high-density three-dimensional shape of the object M Data can be easily obtained. Further, since the plurality of cross sections to be set are set in the same reference direction with respect to the object M, the two-dimensional coordinate data obtained for each cross section can be combined (two-dimensional coordinates without considering the direction of each cross section). Since a set of data can be generated and used, two-dimensional coordinate data can be easily obtained.

Sampling window Here, the approximate curve that approximates the shape of the surface of the object M included in the cross section is not only the three-dimensional coordinate data of the point of the position where the cross section is set, but also the reference direction of the position where the cross section is set An approximate curve that approximates the shape of the surface of the object M at the position where the cross section is set may be specified in addition to the three-dimensional coordinate data of the points existing in the sampling window provided in the front-rear direction.

  FIG. 17 is a schematic diagram showing the relationship between the position where the cross section is set and the sampling window. This figure shows that the cross sections CSn-1, CSn, CSn + 1 are set to X coordinates Xn-1, Xn, Xn + 1. Although not shown in the figure, cross sections parallel to the range of X <Xn−1 and X> Xn + 1 are also arranged at intervals p. This figure also shows that the sampling window for the cross section CSn having a width W represented by a hatched rectangle is set in the range of Xn−W / 2 to Xn + W / 2. Similarly, sampling windows having the same length W are set for the other cross sections. The length W of the sampling window is stored in the storage unit 16 in advance. In the figure, a point drawn with a black circle means a point existing in the sampling window, and a point drawn with a white circle means a point outside the sampling window. In the illustrated example, there are three cross sections in the sampling window. However, the length W of the sampling window and the interval p between the cross sections are determined independently, and W must be an integral multiple of p. There is no.

  The apparatus 1 uses the X coordinate at which the cross section is set and the length W of the sampling window, and the starting point of the X coordinate of the sampling window set for each cross section is Xn−W / 2, and the end point of the X coordinate. Is determined as Xn + W / 2. That is, the apparatus 1 sets the sampling window including the position Xn where the cross section is set so that the intermediate position in the reference direction of the sampling window matches the position Xn where the cross section is set. In addition, the apparatus 1 uses the three-dimensional coordinate data in the range from the start point to the end point of the sampling window among the three-dimensional coordinate data stored in the DB 1, and uses the object M included in the cross section CSn. Specify an approximate curve that approximates the surface of.

  As described above, in the present invention, the section CSn is set as a point on the surface of the object M included in the cross section CSn, which is used when specifying an approximate curve that approximates the shape of the surface of the object M included in the cross section CSn. In addition to the set point on the X coordinate Xn, the point in the sampling window W corresponding to the X coordinate Xn where the cross section CSn is set is included.

  Thus, when an approximate curve is specified by providing a sampling window, the amount of information (number of points on the surface of the object M) used to specify the approximate curve is 3 of the X coordinate where the cross section is set. Since it increases compared to the case where only the dimensional coordinate data is used, an approximate curve that approximates the shape of the surface of the object M more accurately can be obtained. As a result, the amount of change in the reference direction of the approximate curve set for each cross section (for example, when a circle is specified as the approximate curve, the amount of change in the radius of the circle for each adjacent cross section) can be reduced.

  FIG. 18 is a schematic diagram illustrating an example of an approximate curve specified for each adjacent cross section. FIG. 6A shows a state in which the shapes of the objects in the respective imaging regions Ar0, Ar1, Ar2, Ar3 calculated based on the distance images acquired by imaging the objects by the four imaging devices are combined. Yes. In the figure, X0 and X0 + 1 indicate the setting positions of adjacent cross sections (the paper plane, the horizontal direction is the reference direction). This figure shows that the shape of the surface of the object is discontinuous between the position X0 where the cross section is set and the position X0 + 1 where the cross section is set due to the difference in the field of view of the imaging regions Ar1 and Ar3. Yes.

  FIG. 4B shows an example of an approximate curve at the position X0 on the left side and an example of an approximate curve at the position X0 + 1 on the right side of the figure. The figure shows that even if some of the points on the surface included in the adjacent cross-sections are at the same position, the center position and radius of the circle as the approximate curve are different.

  Even when adjacent approximate curves change greatly as described above, as described above, by specifying one approximate curve based on the points in the sampling window, the change in the approximate curve of the adjacent cross section The amount can be reduced. As a result, the value of the direction orthogonal to the reference direction of the obtained two-dimensional coordinate data is smoothly displaced along the reference direction. For example, the shape of the surface of the object M is steep due to the difference in the field of view of the imaging region. The change can be converted into a smooth change on the projected image.

  FIG. 19 is a schematic diagram illustrating another example of the approximate curve specified for each adjacent cross section. FIG. 18A is the same as FIG. In the figure, the four lines between the lines indicating the cross-section positions X0 and X0 + 1 (paper, vertical line) indicate the positions of corresponding points on the surface of the object included in the adjacent cross-section. ing.

  FIG. 4B shows an example of an approximate curve at the position X0 on the left side and an example of an approximate curve at the position X0 + 1 on the right side of the figure. In this figure, by specifying one approximate curve based on points on the surface of a plurality of adjacent cross sections, the change in the approximate curve between adjacent cross sections becomes gentle, that is, as the approximate curve on the left in the figure. This shows that the radius of the circle increases, the radius of the circle as the approximate curve on the right in the figure decreases, and as a result, the difference in the radius of the circle as the approximate curve of adjacent cross sections decreases.

  As a result, as shown in FIG. 8C, the change amount of the position of the corresponding point on the surface of the object included in the adjacent cross section is smaller than that in FIG. Yes.

  Here, a line connecting the centers of the circles as an approximate curve specified for each of a plurality of adjacent cross sections is a non-straight line. As described above, the line connecting the centers of the circles as the approximate curve specified for each cross section is also a non-linear line. That is, in the present invention, each circle as an approximate curve (an approximate curve specified for each of a plurality of cross sections or an approximate curve specified for each cross section) specified corresponding to a plurality of set cross sections. The line connecting the centers of is non-linear.

Change in pitch (P) In order to obtain two-dimensional coordinates that approximate the three-dimensional shape of the object M with high accuracy, it is desirable to shorten the pitch between the set cross sections. However, when this pitch is shortened, the number of cross sections to be set increases, and the amount of information becomes enormous.

  Therefore, in the present invention, the initial value of the pitch is determined (stored in the storage unit 16), the approximate curve of each cross section calculated with the initial value of the pitch is evaluated, and the pitch is determined depending on the evaluation result. The approximate curve may be specified by resetting (for example, setting shorter than the initial value) and resetting the cross section using the pitch after resetting (after change). That is, for example, when a cross-section is set with an initial value and a circle as an approximate curve of each cross-section is specified, a difference in radius between adjacent circles is a predetermined difference (threshold: stored in the storage unit 16 in advance). When it is larger, there is a possibility that the size of the object M is rapidly changing in the range of the reference direction in which the adjacent cross sections are set. Therefore, in such a case, the pitch is narrower than the initial value, the cross-section is set again, the approximate curve is specified, and the two-dimensional coordinates are calculated to approximate the three-dimensional coordinates with high accuracy. Two-dimensional coordinates can be obtained. Note that the pitch change amount (for example, the pitch is changed to 60% of the current pitch) is stored in the storage unit 16 in advance.

● Summary of the features of the device The features of the device described so far are listed below.

(Feature 1)
A projection image generation device that generates two-dimensional coordinate data of a projected image obtained by projecting the surface onto a two-dimensional plane from three-dimensional coordinate data of a distance image representing a three-dimensional shape of the surface of the object,
An acquisition unit for acquiring the three-dimensional coordinate data;
A cross-section setting unit for setting a plurality of cross-sections parallel to each other at a predetermined pitch in a specific direction of the object;
An approximate curve specifying unit that specifies an approximate curve that approximates the shape of the surface included in the cross section using the three-dimensional coordinate data; and
For each point on the surface included in the cross section, a two-dimensional coordinate data calculation unit that calculates the two-dimensional coordinate data of the point based on a distance from a reference point on the approximate curve;
A storage unit for storing the two-dimensional coordinate data;
Having
A projected image generation apparatus characterized by that.

(Feature 2)
The approximate curve specifying unit specifies an approximate curve that approximates the shape of the surface included in the cross section using the three-dimensional coordinate data of a point in a sampling window including a position where the cross section is set.
A projected image generation apparatus according to Feature 1.

(Feature 3)
The sampling window including the position where the cross section is set is set so that the middle position of the sampling window in the specific direction matches the position where the cross section is set.
A projected image generation apparatus according to Feature 2.

(Feature 4)
The approximate curve is a circular arc,
The two-dimensional coordinate data calculation unit calculates two-dimensional coordinate data for each point on the surface included in the cross section,
The radius of the circle;
An angle formed by a line connecting the center of the circle and the reference point, and a line connecting the center of the circle and the point;
Based on
The projected image generation apparatus according to any one of features 1 to 3.

(Feature 5)
A line connecting the centers of the respective circles as the approximate curve specified corresponding to the plurality of set cross sections is a non-straight line,
A projected image generation apparatus according to Feature 4.

(Feature 6)
A pitch changing unit that changes the length of the pitch based on the radius of the circle as the approximate curve,
The cross-section setting unit sets a plurality of cross-sections parallel to each other based on the changed pitch,
The approximate curve specifying unit specifies a new approximate curve that approximates the surface included in a new cross section set based on the changed pitch,
The two-dimensional coordinate data calculation unit calculates, for each point on the surface included in the new section, two-dimensional coordinate data of the point based on a distance from a reference point on the new approximate curve.
The projected image generation device according to Feature 4 or 5.

(Feature 7)
A specific direction determining unit that determines the specific direction based on the three-dimensional coordinate data;
Comprising
The projected image generation device according to any one of features 1 to 6.

(Feature 8)
The specific direction determination unit determines the specific direction based on a distribution of point groups on the surface of the object calculated from the three-dimensional coordinate data.
The projection image generation apparatus according to Feature 7.

(Feature 9)
The specific direction determining unit distributes a normal vector of each point on the surface of the object on a unit sphere, and a plane passing through the origin of the unit sphere with respect to a point group at an end point of the normal vector of each point. Fit, determine the direction of the normal of the plane as the specific direction,
The projection image generation apparatus according to Feature 8.

(Feature 10)
The specific direction determining unit
A normal vector of each point on the surface of the object is distributed on a unit sphere, and a plane passing through the origin of the unit sphere is applied to a point group of an end point of the normal vector of each point, and the normal of the plane Is specified as the axial direction of the initial basic axis,
Determining one direction as the specific direction among a plurality of directions in which the axial direction of the initial basic axis is deviated by a predetermined angle in a different direction;
The projection image generation apparatus according to Feature 8.

(Feature 11)
The specific direction determining unit
Based on the distribution of point groups on the surface of the object calculated from the three-dimensional coordinate data, specify the axial direction of the initial basic axis,
Determining one direction as the specific direction among a plurality of directions in which the axial direction of the initial basic axis is deviated by a predetermined angle in a different direction;
The projection image generation apparatus according to Feature 7.

M Object Cam Imaging device A, Ar Imaging range SP Initial basic axis SA Basic axis candidate 1
SB basic axis candidate 2
p Distance between sections (pitch)
CS Cross section C Center of circle as approximate curve r Radius of circle as approximate curve W Sampling window length

Claims (13)

A projection image generation device that generates two-dimensional coordinate data of a projected image obtained by projecting the surface onto a two-dimensional plane from three-dimensional coordinate data of a distance image representing a three-dimensional shape of the surface of the object,
An acquisition unit for acquiring the three-dimensional coordinate data;
A cross-section setting unit for setting a plurality of cross-sections parallel to each other at a predetermined pitch in a specific direction of the object;
An approximate curve specifying unit that specifies an approximate curve that approximates the shape of the surface included in the cross section using the three-dimensional coordinate data; and
For each point on the surface included in the cross section, a two-dimensional coordinate data calculation unit that calculates the two-dimensional coordinate data of the point based on a distance from a reference point on the approximate curve;
A storage unit for storing the two-dimensional coordinate data;
Having
A projected image generation apparatus characterized by that. The approximate curve specifying unit specifies an approximate curve that approximates the shape of the surface included in the cross section using the three-dimensional coordinate data of a point in a sampling window including a position where the cross section is set.
The projected image generation apparatus according to claim 1. The sampling window including the position where the cross section is set is set so that the middle position of the sampling window in the specific direction matches the position where the cross section is set.
The projected image generation apparatus according to claim 2. The approximate curve is a circular arc,
The two-dimensional coordinate data calculation unit calculates two-dimensional coordinate data for each point on the surface included in the cross section,
The radius of the circle;
An angle formed by a line connecting the center of the circle and the reference point, and a line connecting the center of the circle and the point;
Based on
The projected image generation apparatus according to claim 1. A line connecting the centers of the respective circles as the approximate curve specified corresponding to the plurality of set cross sections is a non-straight line,
The projection image generation apparatus according to claim 4. A pitch changing unit that changes the length of the pitch based on the radius of the circle as the approximate curve,
The cross-section setting unit sets a plurality of cross-sections parallel to each other based on the changed pitch,
The approximate curve specifying unit specifies a new approximate curve that approximates the surface included in a new cross section set based on the changed pitch,
The two-dimensional coordinate data calculation unit calculates, for each point on the surface included in the new section, two-dimensional coordinate data of the point based on a distance from a reference point on the new approximate curve.
6. The projected image generation apparatus according to claim 4 or 5. A specific direction determining unit that determines the specific direction based on the three-dimensional coordinate data;
Comprising
The projection image generation apparatus according to claim 1. The specific direction determination unit determines the specific direction based on a distribution of point groups on the surface of the object calculated from the three-dimensional coordinate data.
The projection image generation apparatus according to claim 7. The specific direction determining unit distributes a normal vector of each point on the surface of the object on a unit sphere, and a plane passing through the origin of the unit sphere with respect to a point group at an end point of the normal vector of each point. Fit, determine the direction of the normal of the plane as the specific direction,
The projection image generation apparatus according to claim 8. The specific direction determining unit
A normal vector of each point on the surface of the object is distributed on a unit sphere, and a plane passing through the origin of the unit sphere is applied to a point group of an end point of the normal vector of each point, and the normal of the plane Is specified as the axial direction of the initial basic axis,
Determining one direction as the specific direction among a plurality of directions in which the axial direction of the initial basic axis is deviated by a predetermined angle in a different direction;
The projection image generation apparatus according to claim 8. The specific direction determining unit
Based on the distribution of point groups on the surface of the object calculated from the three-dimensional coordinate data, specify the axial direction of the initial basic axis,
Determining one direction as the specific direction among a plurality of directions in which the axial direction of the initial basic axis is deviated by a predetermined angle in a different direction;
The projection image generation apparatus according to claim 7.   A projection image generation program which causes a computer to function as the projection image generation apparatus according to any one of claims 1 to 11. This is a projected image generation method executed by a device that generates two-dimensional coordinate data of a projected image obtained by projecting the surface onto a two-dimensional plane from three-dimensional coordinate data of a distance image representing the three-dimensional shape of the surface of the object. And
The device is
Obtaining the three-dimensional coordinate data;
Setting a plurality of parallel cross sections separated by a predetermined pitch in a specific direction of the object;
Identifying an approximate curve that approximates the shape of the surface included in the cross-section using the three-dimensional coordinate data;
For each point on the surface included in the cross section, calculating the two-dimensional coordinate data of the point based on a distance from a reference point on the approximate curve;
Storing the two-dimensional coordinate data;
Having
A projected image generation method characterized by the above.