CN1099645C - Surface transfer method for point group data - Google Patents

Abstract

The invention discloses a curved surface transformation method for point group data. The method obtains curved surface data (S) from the actually measured point group data (P) of the three-dimensional coordinates of the curved surface of the measured object (10) measured by a three-dimensional measuring instrument. It is that, for a three-dimensional coordinate system, after setting a basic shape surface (5a) with a prescribed graphic shape near the distribution surface (10a) of (P) or at the boundary position, multiple reference points (Q ), set the directional vector (6a) indicating the direction to (10a) on (Q), and then calculate the process of crossing the extension line of (6a) in the indicated direction and (10a) for each reference point (Q) Point group data (D), and obtain (S) according to (D).

Description

Surface Transformation Method for Point Group Data

technical field

The invention relates to a curved surface transformation method for obtaining the curved surface data from the actually measured point group data of the three-dimensional coordinates of the measured object measured by a three-dimensional measuring instrument.

Background technique

In the past, the known methods include the spline method and the NURBS method, etc., which are used to measure the three-dimensional coordinates (x, y, z) data of the measured object from the three-dimensional measuring instrument (hereinafter referred to as the measured point group data), The curved surface data of the measured object curved surface is obtained.

The three-dimensional measuring instrument shown in Figure 14 uses the laser light 12 emitted from the direction shown in Figure A to obtain the point coordinates P1 (x1, y1, z1) of each light point from the reflected light from the curved surface 10a of the measured object 10. , P2(x2, y2, z2), . . . , Pi(x _i , y _i , z _i ). These P1(x1, y1, z1)-Pi(x _i , y _i , zi ₎ become the actually measured point cloud data Pa(x, y, z) of the surface shown in 10a in the figure. Then, any one of the object to be measured 10 or the laser scanner 11 is rotated by 90°, and the actual measurement point cloud data Pb(x, y, z) and sample the measured point group data Pc(x, y, z) and P _D (x, y, z) sequentially from the directions shown by C and D in the figure. After sampling the entire surface of the measured object 10 as described above, three-dimensional measured point cloud data P(x, y, z) is obtained.

In this way, the obtained three-dimensional measured point group data is a collection of multiple points. According to the collection of these points, in the data sampled from each direction (such as A direction and B direction), the operator organizes and selects the repeated data and useless data of the boundary line, and only processes the sorted points Group data, use a mouse, etc., to perform three-dimensional drawing on a computer monitor.

Here, the computer obtains curved surface data from the processed point cloud data drawn on the display using a method such as NURBS.

Such curved surface data can be used, for example, for three-dimensional shape object processing by a three-dimensional modeling machine.

In the previous method for obtaining curved surface data, the 3D measured point cloud data sampled from the 3D measuring instrument is a collection of multiple points, and the points obtained in each direction must be processed into point cloud data that can be processed. Operations are sorted and selected issues.

In other words, points obtained from various directions do not necessarily have the same data at their borders, etc., and in order to display a three-dimensional point group, it is necessary to sort out duplicate data and useless data. For this reason, an input operation and a drawing operation of a computer are required, and a large number of operation man-hours are required. At the same time, there is a problem that the surface data depends on the operator because input errors, selection errors, drawing errors, and loss of useful data occur when processing point group data.

Therefore, a method for easily obtaining a curved surface from unorganized actual measurement point group data obtained by a three-dimensional measuring instrument without manual labor is desired.

Contents of the invention

In order to solve the problems in the aforementioned prior art examples, the curved surface transformation method of the present invention obtains the curved surface data from the measured point group data of the three-dimensional coordinates of the curved surface of the measured object measured by the three-dimensional measuring instrument, and is characterized in that, for the three-dimensional coordinate system, After setting a basic shape surface with a prescribed graphic shape near the distribution surface of the measured point group data or at the boundary position, determine a plurality of reference points on the basic shape surface, and point to the distribution surface of the measured point group data on each reference point Set the directional vector indicating the direction, then obtain the processing point cloud data that the extension line of the directional vector in the shown direction crosses the distribution plane of the measured point cloud data for each reference point of the basic shape surface, and according to the Process point group data to obtain surface data.

In addition, it is preferable to use a point intersecting a microplane created by using the measured point cloud data of the neighborhood extended in the direction indicated by the directional vector as the processed point cloud data.

The present invention is constituted as described above, and the following effects can be obtained. That is to say, adopt the present invention, for a plurality of reference points on the basic shape plane with the prescribed graphic shape set on the three-dimensional coordinates, the extension of the direction shown by the directed vector to the direction indicated on the distribution plane of the measured point group data can be utilized The line intersects with the measured point group data distribution plane to obtain the processed point group data. Specifically, a plurality of basic shape surfaces of various figures are prepared in advance, and one of them is selected corresponding to the curved surface of the object, and a plurality of reference points are set on the basic shape surface, and these reference points are moved to The point where the measured point group data distribution plane crosses, regardless of the repetition and inconsistency in the measured and processed point group data, can obtain the ordered point group data required when using the NURBS method from the basic point group on the basic shape surface (processing point group data), and can easily calculate the curved surface of the three-dimensional surface.

In addition, if the point that intersects with the tiny surface created by the measured point cloud data in the vicinity of the direction indicated by the directed vector is used as the processed point cloud data, then the point cloud data at the adjacent position indicated by the directed vector is used to create It is possible to set hypothetical points on the distribution surface shown by the directed vector, even for the unorganized measured point group data that has errors and defects due to incompleteness, it can also obtain the processing point group that can be processed. data.

According to the present invention, it is possible to provide a method for obtaining curved surface data from actually measured point cloud data in a very short time by reducing the conventionally necessary operation of manually selecting data required for curved surface transformation from actually measured point cloud data.

Description of drawings

Fig. 1 shows a block diagram of a first embodiment of the present invention.

Figure 2 shows a flow chart of the method.

Fig. 3 shows a perspective view of a measurement object.

Fig. 4 is an explanatory view showing a basic shape surface.

Fig. 5 is an explanatory diagram of a directed vector.

FIG. 6 is an explanatory diagram showing movement of points.

Fig. 7 shows a perspective view of the upper curved surface.

Fig. 8 is a perspective view showing methods of determining other surfaces.

Fig. 9 shows a configuration diagram of a second embodiment of the present invention.

Fig. 10 is an explanatory view showing the micro-faces.

Figure 11 shows a flowchart of the method.

Fig. 12 shows a perspective view of another application example.

Fig. 13 is a perspective view showing an output example thereof.

FIG. 14 is an explanatory diagram showing a sampling method of actually measured point cloud data as an example of conventional technology.

Detailed ways

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

Example 1

1 to 8 show a first embodiment of the present invention.

The arithmetic device 1 shown in FIG. 1 includes: an actual measured point cloud data input unit 2 for inputting each coordinate (x, y, z) of the actual measured point cloud data P of the free-form surface measurement object 10 measured by the three-dimensional measuring instrument 7, including Have the basic shape surface 5a of pre-registered rectangular plane, cylindrical surface, spherical surface, rectangular cylinder, square cylinder and other prescribed graphic shapes and the memory data 9 of the directed vector 6a representing its direction, which is input by the measured point group data The actual measured point cloud data P and the memory data 9 of the unit 2 convert the actual measured point cloud data P into the arithmetic processing unit of the curved surface data S, and the curved surface data output unit 4 stores the calculated curved surface data S and outputs it to the three-dimensional modeling processing machine 8 .

FIG. 2 shows its processing flow. In step 1, the actually measured point cloud data P measured by the three-dimensional measuring instrument 7 is loaded into the actually measured point cloud data input unit 2 . P1-P33 shown in FIG. 3 represent the positions of the actually measured point cloud data P of the upper surface portion 10a of the telephone handset 10 which is the object to be measured. In addition, in FIG. 3 , although 33 actually measured point group data are shown on the drawing, actually measured point group data P may take tens of thousands of data.

Here, the operator selects the basic shape surface 5a suitable for the upper surface portion 10a shape of the telephone receiver 10 from the memory data 9, and sets the rectangular plane shown in 5a in Fig. 4 to the adjacent three-dimensional space of the actually measured point cloud data P (step 2).

In step 3, for each reference point cloud data Q on the rectangular plane 5a, a directional vector 6a is set so as to be orthogonal to the distribution plane direction of the measured point cloud data P.

As described above, the memory data 9 in which the rectangular plane 5 a and the directional vector 6 a are set are input to the arithmetic processing unit 3 together with the measured point data P of the measured point group data input unit 2 .

In the calculation processing unit 3, each reference point cloud data Q in the rectangular plane 5a is moved in the direction indicated by the directional vector 6a, and the processed point cloud data D intersected with the measured point cloud data P is calculated on the distribution plane 10a.

In the first embodiment, regardless of the repetition and inconsistency of the measured point group data P, the multiple measured points on the distribution plane 10a are only based on the reference points Q ₁ -Q ₃₃ in the rectangular plane 5a shown in FIG. 6 . From the point cloud data P, the actually measured point cloud data P required for conversion into the curved surfaces shown in P1-P33 in the figure is obtained.

That is to say, the rectangular plane 5a in Fig. 6 is divided into 11 parts in the horizontal direction and each reference point Q1-Q33 on each position after being divided into 2 parts in the vertical direction, according to the directed vector 6a, each vertex Q1, Q3, Q31, Q33 respectively intersect with vertices P1, P3, P31, P33 in the distribution plane 10a. Points Q2 and Q32 on the side intersect P2 and P32, Q6, Q9...Q30 intersect P6, P9...P30, and Q4, Q7...Q28 intersect P4, P7...P28. Then, interior points Q5, Q8...Q29 intersect with P5, P8...P29.

As mentioned above, only the calculated intersection position is calculated as the processing point cloud data Di that can be processed on the curved surface, and the above-mentioned reference points Q in the rectangular plane 5a are then moved to the processing point cloud data Di position in the distribution surface 10a, Thus, the curved surface Sa shown in FIG. 7 is generated (step 5). That is to say, the arithmetic processing unit 3 uses the NURBS method to obtain the three-dimensional surface data (ie curved surface data) of the distribution surface 10a according to the aforementioned processed point cloud data Di.

After the distribution surface 10 a shown in FIG. 7 is subjected to the aforementioned arithmetic processing, the measured point cloud data P is converted into curved surface data Sa.

As described above, the curved surface data Sa representing the top surface of the handset 10 is temporarily stored in the curved surface data output unit 4 .

Next, for the bottom surface 10b of the telephone handset 10 shown in FIG. 8, a rectangular plane 5b is used as the basic shape surface, and a direction 6b orthogonal to the distribution surface 10b is used for the directional vector. Similar to the above-mentioned upper surface 10a, the processed point cloud data D is obtained only for the required measured data P intersecting with each reference point cloud data Q on the rectangular plane 5b, and the curved surface data Sb is calculated from the processed point cloud data D.

Furthermore, at the two end surfaces 10c, 10d of the telephone receiver 10, respectively adopt the cylindrical surfaces 5c, 5d with partial defects and the directional vectors shown in 6c, 6d in Fig. The curved surface data Sd, Sc and Sd of each part are obtained and stored in the curved surface output unit 4 in sequence. When the aforementioned process is repeated to obtain curved surface data for all planes, the curved surface data output unit 4 outputs the curved surface data S to the three-dimensional modeling machine 8 .

As mentioned above, adopting said embodiment 1, for the measured point cloud data P of the measuring object 10 measured by the three-dimensional measuring instrument 7, only the basic shape surface 5a and its directional vector 6a are provided to obtain the required By processing the point cloud data D and moving each reference point Q, the curved surface data S of the measurement object 10 can be obtained.

Example 2

Embodiment 2 of the present invention will be described below with reference to FIGS. 9-13 .

Embodiment 2 is to the calculation method of the point of reference point Q in embodiment 1 and required measured point group data P intersecting, its feature is to set hypothetical tiny plane B on the distribution plane of measured and group data P, and calculate and tiny The point where face B intersects. That is to say, it is not necessary to make the reference point Qi shown in FIG. 6 coincide with the actually measured point cloud data Pi, and it is also possible to use The microplane B made by the point group adjacent to the intersection obtains the intersection position.

Therefore, other structures of Embodiment 2 are partly the same as in Embodiment 1, so in FIGS. 9-13 , the same symbols are attached to the same parts and their detailed descriptions are omitted.

The flow shown in FIG. 11 is the same as in the first embodiment in steps 1 to 3, and the basic shape plane 5a and its directional vector 6a are set for the measured point cloud data P.

Next, in Embodiment 2, a processing method in a case where the reference point Qi on the rectangular plane 5a shown in FIG. 9 does not necessarily correspond to Pi on each distribution surface 10a will be described.

Step 4a-Step 4b shown in FIG. 11 is a part related to the formation of the microfacet B. FIG. Here, the point Qi shown in FIG. 9 does not have an intersection point Pi that intersects the actual measurement point cloud data P in the direction of the directional vector 6a. Then, take the points near the position extending from the reference point Qi to the direction of the directional vector 6a and intersect with the distribution 10a (points P1 and P2 in Figure 10), and calculate the micro-plane B assumed based on this point, to obtain a processable Point Di (processing point group data D).

Then, in step 5a, vertices and side points and internal points of the basic shape surface 5a that are unclear where to move are moved to the positions of the intersecting micro-surfaces B to generate a curved surface.

The Di point shown in FIG. 10 shows how the Qi point has moved onto the tiny plane B formed by the P1 and P2 points.

As mentioned above, according to the above-mentioned second embodiment, even when it is not clear where to move, the moving arrival point can be specified from the surrounding points with certainty, and the curved surface data S can be obtained by means of NURBS or the like. That is to say, in the multi-point collection of the measured point group data P, even if there are data groups that have not been sorted out, such as repeated and useless data or defects, it can be adapted from the adjacent data to make a small surface B, so as to accurately obtain It becomes the processed point cloud data D at the point where the move can be processed, and the curved surface S is obtained.

Fig. 12 is the case where the present invention is applied to a human ear hole model. The ear hole model is made of silicon material from the side of the ear that cannot move freely, and the actual measured point cloud data of repeated 45° scanning measurements in the hole axis direction with a three-dimensional measuring instrument P, adopt the basic shape surface 5a of the cylindrical surface shape, divide the cylindrical surface into a grid, and set the directional vector 6a for each reference point Qi toward the central axis. In this case, it takes more than 3 hours to obtain the data that can be handled by manual manual operation conventionally by representing the measured point cloud data with a total of about 50,000 points, and to generate a curved surface. If the curved surface conversion method of the measured point group data of the present invention is adopted, the intersection point is obtained by making the micro-surface B from the surrounding points, so basically it does not depend on the number and distribution shape of the measured point group data P, and is compatible with defects and errors. Alternatively, irrespective of the repetition, the actually measured point cloud data P of about 50,000 points can be converted into the curved surface data S shown in FIG. 13 in about one minute. The curved surface data S of such an ear canal can be used as processing data for a three-dimensional modeling machine when custom-made hearing aids are manufactured.

In addition, in the foregoing embodiments, the minute plane B is not limited to a rectangular plane, and may be a triangular plane, a circular plane, and an arbitrary curved surface. In addition, the minute surface B may be a surface created using a coordinate point group obtained by averaging some adjacent coordinate points. The size of the tiny surface B can be fixed, or it can be obtained according to the variation of the measured point group data. In addition, although the arithmetic processing unit 3 shows a method of obtaining a curved surface using the NURBS method, it is not limited to the NURBS method, and may be obtained by other methods such as the spline method. That is to say, the present invention is not limited to the foregoing embodiments, as long as it is based on the true spirit of the present invention, various modifications are possible, and are included in the scope of all patent claims.

Claims (2)

1. A curved surface transformation method of point group data, this curved surface transformation method obtains curved surface data from the measured point group data of the curved surface three-dimensional coordinates of the measured object measured by a three-dimensional measuring instrument, and is characterized in that the three-dimensional coordinate system After setting a basic shape surface with a prescribed graphic shape near or on the boundary of the distribution surface of the measured point cloud data, a plurality of reference points are determined on the basic shape surface, and the distribution surface of the measured point cloud data is drawn on each reference point. Set the directional vector indicating the direction, then obtain the processing point cloud data that the extension line of the directional vector in the shown direction crosses the distribution plane of the measured point cloud data for each reference point of the basic shape surface, and according to the Process point group data to obtain surface data. 2. the curved surface conversion method of point cloud data as claimed in claim 1 is characterized in that, with the point that with the micro-face intersecting that the adjacent place measured point cloud data after extending toward the direction shown in directed vector is made, as processing point group data.

Images