JPH0697166B2 - 3D data correction method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a three-dimensional data correction method for point cloud data in a three-dimensional space obtained by a three-dimensional position measuring machine or the like.

(Prior Art) A contact point of a model shape can be obtained by bringing a probe such as a three-dimensional position measuring machine into contact with the model surface with one axis fixed and scanning the model surface a plurality of times at predetermined intervals. it can. For example, as shown in FIG. 7, the X-axis is fixed, the scanning (measuring) direction is the Y-axis direction, the probe ST is scanned on the surface of the model M five times at the measurement interval a, and predetermined interpolation is performed.
It is possible to obtain point cloud data indicated by black circles in the drawing. In this point cloud data, each point is connected as a series of fine straight line sections in the measuring direction (S ₁ , S ₂ , ..., S ₅ shown in the figure). As a method of using such point cloud data, processing information may be used as it is, or shape feature recognition that extracts only the corners and valleys of the measured shape (for example, the corner MC of the model M shown in FIG. Extraction as shown in FIG. 8 (solid line portion) or shape operation for expanding the end portion of the measured shape along the shape (for example, the end portion ME of the model M shown in FIG. 7 is expanded as shown in FIG. 9 (solid line portion). Part)) etc.

(Problems to be Solved by the Invention) In order to improve the accuracy of point cloud data obtained by measuring with a probe such as the above-described three-dimensional position measuring instrument, it is necessary to completely fix one axis of the probe. is there. However, it is impossible to completely fix it physically, and it becomes point cloud data including a fluctuation component. Then, by using the point cloud data including such a variation component (for example, FIG. 10), the position to be recognized is erroneously extracted when performing the above-described shape feature recognition (for example, the solid line portion shown in FIG. 11), In addition, there is a problem that when the shape operation is performed, the shape is expanded in a random direction (for example, a solid line portion shown in FIG. 12).

The present invention has been made under the circumstances as described above, and an object of the present invention is to perform accurate shape feature recognition and stable shape operation even if the point cloud data of the model shape includes a fluctuation component. The object of the present invention is to provide a three-dimensional data correction method capable of performing the above.

(Means for Solving the Problems) The present invention relates to a method for correcting three-dimensional data in point cloud data in a three-dimensional space obtained by a three-dimensional position measuring machine or the like. 3 obtained by scanning the model surface multiple times at a predetermined interval
The point cloud data of the dimensional space is input, and connecting lines with points on the line connecting each point of one point cloud data of one point scan data of one adjacent scan to each point of the other point cloud data, respectively. Obtained, to find the surface containing the ideal scanning line at the time of measurement by the tracing stylus, to find the intersection of the line and the connecting line and the surface connecting the points of the point cloud data in the scanning direction, This is achieved by a feature that the obtained intersection is used as corrected point cloud data.

(Operation) According to the three-dimensional data correction method of the present invention, an intersection between a connection line stretched like a net between point cloud data and a plane including an ideal scanning line is obtained, and the obtained intersection is used as the correction point cloud data. Therefore, the fluctuation component included in the original point cloud data can be completely removed.

(Embodiment) FIG. 1 is a block diagram showing an example of an apparatus that realizes the three-dimensional data correction method of the present invention. Point cloud data measured by a three-dimensional position measuring machine or the like according to a command from the operation panel 1. Is input and the point cloud data S _{i for} one scan (i = 1, 2, ... N) is sequentially stored in the measurement data reading memory 3, and the point cloud data S _{i−1 for} one adjacent scan is input. , S _i Find the point on the other point cloud data that has the shortest distance from each point of one point cloud data to the line connecting each point of the other point cloud data, and connect the points with the shortest distance , The connection information CT _m (m−1,2, ..., N−1) consisting of the respective connection lines Cik (k = 1,2, ...) Is obtained and sequentially stored in the connection information memory 4, and at the time of measurement. The plane containing the ideal scanning line of the point cloud is obtained, and the connection of the point cloud data (the measurement direction connection and the connection line Sequentially obtain the intersection, and a processing device 2 to be output to the outside as the point cloud data obtained by correcting the obtained intersecting point.

In such a configuration, its operation will be described with an example of the flow chart of FIG. 2 and the point cloud data of FIG. 3. When the processing device 2 receives a command from the operation panel 1, it is measured by a three-dimensional position measuring machine or the like. The point cloud data is read and the point cloud data for one scan is sequentially stored in the measurement data reading memory 3 (step S1). Then, it is confirmed whether or not the storage of the point cloud data is completed (step S2). If the storage of the point cloud data is not completed, the process returns to step S1 and the above-described operation is repeated. On the other hand, when the storage of the point cloud data S _i-1 and S _i is completed in the determination step S2, one of the point cloud data of adjacent one scan, for example, S.
_From the point Q ₃ (P ₃ ) on _i-1 (S _i ) to the other point cloud data S
_{i (S i-1)} the other of the point group data _{S i (S i-1)} points on P ₃ which distance is the shortest up '(Q _3') to seek binding line C _i3 a (C _i4) The obtained operation is performed for all points, and the joint information CT _i composed of the obtained joint lines C _{i1 to} C _i5 is stored in the joint information memory 4. It should be noted that points that are clearly recognized as corners of the shape (for example, P ₂ and Q ₂ ), start point of the point cloud data (for example, P ₁ and Q ₁ ), end point (for example, P ₄ and Q ₂ ).
The connection lines in Q ₄ ) are C _i1 , C _i2 and C _i5 which are obtained by connecting the points without obtaining the shortest distance. Then, this operation is sequentially performed from the point cloud data for the first scan (step
S3 to S5), it is recognized whether or not the point cloud data for the final (N) th scan has been completed (step S6), and if the point cloud data for the Nth scan has not been completed, step S4. Then, the above operation is repeated. On the other hand, in the decision step S6, when it is completed up to N-th point cloud data of the scanning operation, for example obtains the plane PL _i that contains the ideal scan line when the S _i th measurement, the plane PL Intersection R ₁ between _i and the connection of the point cloud data stretched in a mesh
Seeking R _8, the operation of outputting to the outside as the intersection R ₁ to R ₈ point cloud data obtained by correcting the determined sequentially performed from the ideal scanning line during the first measurement (step S7 to S11). And
It is confirmed whether or not the ideal scanning line at the time of the final (N) measurement is completed (step S12), and if the ideal scanning line at the time of the Nth measurement is not completed, a step is performed.
After returning to S8, the above-described operation is repeated, and when the ideal scanning line for the Nth measurement is completed, all the processes are completed.

4 to 6 show the case where the correction method according to the present invention is applied to the changed point cloud data shown in FIG. First, as shown in FIG. 4, the connecting line (dotted line) at each point of the point cloud data is obtained and formed into a mesh, and for example, the plane PL ₃ including the ideal scanning line at the time of the third measurement is obtained, The corrected point cloud data can be obtained by obtaining the intersection (black circle shown in FIG. 5) of the plane PL ₃ and each connection. In this way, by correcting the point group data S ₁ ′ to S ₅ ′, it is possible to obtain highly accurate point group data S ₁ ″ to S ₅ ″ as shown in FIG.

(Effect of the Invention) As described above, according to the three-dimensional data correction method of the present invention,
Since it is possible to perform accurate shape feature recognition and stable shape operation, even with point cloud data that contains fluctuation components,
It is possible to stably supply products with high dimensional accuracy and reduce manufacturing costs.

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of an apparatus for realizing the three-dimensional data correction method of the present invention, FIG. 2 is a flow chart for explaining the operation thereof, and FIG. 3 is a description of the three-dimensional data correction method of the present invention. FIG. 4 to FIG. 6 are perspective views showing an example of point cloud data, FIG. 4 to FIG. 6 are perspective views showing a process when the three-dimensional data correction method of the present invention is applied to an example of point cloud data, and FIG. A perspective view for explaining an example of a method for obtaining point cloud data, FIGS. 8 and 9 are perspective views for explaining a method for using the point cloud data, and FIG. 10 shows an example of point cloud data including a fluctuation component. FIGS. 11 and 12 are perspective views showing the case where the usage method shown in FIGS. 8 and 9 is applied to the point cloud data including the fluctuation component, respectively. 1 ... Operation panel, 2 ... Processing device, 3 ... Measurement data reading memory, 4 ... Combined information memory.

Claims (1)

[Claims] 1. Point cloud data of a three-dimensional space obtained by scanning a probe on a model surface a plurality of times at predetermined intervals, and inputting one point cloud of the point cloud data for one adjacent scan. From each point of the data, the connecting line with the point on the line connecting the respective points of the other point cloud data is respectively obtained, and the surface including the ideal scanning line at the time of measurement by the above-mentioned probe is obtained, and the point cloud data is obtained. The three-dimensional data correction method is characterized in that a line connecting each point in the scanning direction and an intersection of the connecting line and the surface are obtained, and the obtained intersection is used as corrected point group data.