Arbitrary densification and sparseness processing method for discrete data volume of three-dimensional physical property field

Abstract

The present invention discloses a method for arbitrary densification and sparseness processing of a three-dimensional physical property field discrete data body, comprising the following steps: 1) selecting an observation direction of the original three-dimensional physical property field discrete data body, and constructing a three-dimensional physical property field intermediate data body; 2) interpolating and calculating the value of each discrete data point on each observation surface of the three-dimensional physical property field intermediate data body based on the original three-dimensional physical property field discrete data body; 3) constructing a three-dimensional data body of the three-dimensional physical property field result, the number of layers of this data body in the observation direction is the value after densification and sparseness, and the two dimensions of the two-dimensional observation surface of each layer are the same as those of the intermediate data body; 4) based on the intermediate data body, copying the data points of the top and bottom observation surfaces, and then interpolating and calculating the value of each discrete data point on each observation surface of the three-dimensional data body except the top and bottom surfaces. The present invention has the advantages of clear principle, good applicability, and high processing efficiency.

Description

Technical Field

[0001] The invention relates to a method for arbitrarily densifying and sparsely processing discrete data volumes of three-dimensional physical property fields. Background Art

[0002] The three-dimensional physical property field discrete data volume usually organizes the physical property value data point set in the three-dimensional space in the form of columns, rows, and layers (hereinafter referred to as the three-dimensional data volume). In applications such as the visualization of the three-dimensional physical property field discrete data volume, the columns, rows, and layers correspond to the X, Y, and Z coordinate directions in the Cartesian coordinate system, respectively. In applications, in order to obtain ideal results, it is often necessary to densify or sparse the columns, rows, and layers of the original data. For example, when the distribution of data points in a certain dimension is too sparse, it is necessary to perform densification processing; conversely, when the distribution of data points in a certain dimension is too dense, it is necessary to perform sparse processing.

[0003] When a three-dimensional data body is densified or thinned, if the change of the corresponding dimension before and after processing is an integer multiple, the problem is relatively simple and can be completed directly using linear interpolation or extraction. In reality, it is often difficult to ensure that the change before and after data processing is an integer multiple. It is more necessary to be able to adapt to arbitrary densification or thinning in three dimensions, that is, to support the densification and thinning (arbitrary change pattern) of integer multiples or non-integer multiples of one or more dimensions (arbitrary dimensions and number of dimensions, arbitrary change rate), such as non-integer multiple densification of X dimension, non-integer multiple thinning of Y dimension, non-integer multiple densification of Z dimension, etc. Based on the above analysis, the present invention proposes a processing method for arbitrary densification and thinning of discrete data bodies of three-dimensional physical property fields. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a processing method for arbitrary densification and sparseness of discrete data volumes of three-dimensional physical property fields with clear principles, good applicability and high processing efficiency.

[0005] The technical solution adopted by the present invention is: a method for arbitrary densification and sparseness processing of a three-dimensional physical property field discrete data volume, defining the original three-dimensional physical property field discrete data volume as V 3d_raw , the number of data points in the X, Y, and Z dimensions are N respectively raw_x 、N raw_y 、N raw_z , define the resulting three-dimensional data volume as V 3d_res , the number of data points in the X, Y, and Z dimensions are N respectively res_x 、N res_y 、N res_z ;

[0006] The steps include:

[0007] 1) According to the application characteristics of the physical field, select an observation direction in the original three-dimensional discrete data volume of the physical field, and construct a three-dimensional physical field intermediate data volume V based on the selected observation direction. 3d_inter , three-dimensional physical property field intermediate data volume V 3d_inter In the observation direction, V 3d_raw With the same number of layers, and the intermediate data volume V of the three-dimensional physical property field 3d_inter The positions of each layer and the original three-dimensional discrete data volume V of the physical field 3d_raw The positions of each layer of coincide with each other, but the number of data points in each layer is the number of data points after arbitrary density and sparseness in the two-dimensional directions of the two-dimensional observation surface;

[0008] 2) Based on the original three-dimensional discrete data volume V of the physical field 3d_raw , interpolation calculation of the three-dimensional physical property field intermediate data volume V 3d_inter The value of each discrete data point on each observation surface of ;

[0009] 3) Construct a three-dimensional data volume V of the three-dimensional physical field results 3d_res , the three-dimensional data volume V of the three-dimensional physical field results 3d_res The number of layers in the observation direction is the value after arbitrary density and sparseness. The number of data points in the two dimensions of the two-dimensional observation surface of each layer is the same as the intermediate data volume V of the three-dimensional physical field. 3d_inter The number of data points in the two dimensions of is the same;

[0010] 4) Based on the intermediate data volume V of the three-dimensional physical property field 3d_inter First, copy the data point values ​​of the top and bottom observation surfaces to the three-dimensional data volume V of the three-dimensional physical field result. 3d_res The three-dimensional data volume V of the interpolation calculation result is then obtained on the top and bottom observation surfaces. 3d_res The value of each discrete data point on each observation face except the top and bottom faces.

[0011] In the above-mentioned arbitrary densification and sparseness processing method of the discrete data volume of the three-dimensional physical property field, the specific operation of step 1) is as follows;

[0012] When the observation direction is selected as the Z direction, the observation plane is the XY plane, and the three-dimensional physical property field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. res_x 、N res_y 、N raw_z ;

[0013] When the observation direction is selected as the Y direction, the observation plane is the XZ plane, and the three-dimensional physical property field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. res_x 、N raw_y 、N res_z;

[0014] When the observation direction is selected as the X direction, the observation plane is the YZ plane, and the three-dimensional physical property field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. raw_x 、N res_y 、N res_z .

[0015] In the above-mentioned method for arbitrary densification and sparseness processing of discrete data volumes of three-dimensional physical property fields, in step 2), the following steps are used to calculate the intermediate data volumes V of the three-dimensional physical property fields one by one: 3d_inter The value of each discrete data point on the 1st to Nth observation surfaces, where N is the total number of observation surfaces of the original three-dimensional discrete data volume of the physical property field;

[0016] 2.1) The 3D physical property field intermediate data volume V 3d_inter The i-th observation surface and the original three-dimensional discrete data volume V of the physical field 3d_raw Align the four boundaries of the i-th observation surface, where i∈[1,N];

[0017] 2.2) Using the interpolation method, based on the original three-dimensional discrete data volume V of the physical field 3d_raw The boundary data of the i-th observation surface are used to calculate the intermediate data volume V of the three-dimensional physical property field. 3d_inter The data point values ​​on the four boundaries of the i-th observation surface;

[0018] 2.3) Using the interpolation method, based on the original three-dimensional discrete data volume V of the physical field 3d_raw The observation surface data of the i-th observation surface and the three-dimensional physical property field intermediate data volume V 3d_inter The data point value on the boundary of the i-th observation surface is used to calculate the intermediate data volume V of the three-dimensional physical property field 3d_inter The values ​​of all non-boundary data points in the i-th observation surface.

[0019] In the above-mentioned arbitrary densification and sparseness processing method for the discrete data volume of the three-dimensional physical property field, in step 2.2), the three-dimensional physical property field intermediate data volume V is calculated. 3d_inter When the data point values ​​on the four boundaries of the i-th observation surface are calculated, the data point values ​​on the boundaries of the two dimensional directions of the observation surface are calculated respectively. The calculation method for the data point values ​​on the boundaries of each dimensional direction is as follows:

[0020] When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is the same as the original three-dimensional discrete data volume V of the physical field. 3d_raw At the same time, the original three-dimensional discrete data volume V 3d_rawThe data points on the two boundaries of the dimension direction are directly copied one by one to the intermediate data volume V 3d_inter In the data points corresponding to the corresponding boundaries;

[0021] When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is greater than that of the original three-dimensional discrete data volume V of the physical field. 3d_raw When there are many data points on the boundary of this dimension, it means that the data points on the boundary of this dimension need to be densified. The specific operation is as follows:

[0022] First, according to the original three-dimensional discrete data volume V 3d_raw and the 3D physical property field intermediate data volume V 3d_inter The number of data points N on the boundary of this dimension raw_α and N res_α , calculate the integer multiple densification judgment factor K in the dimension direction d_α , the calculation formula is as follows:

[0023] Then calculate the transformation ratio R α , the calculation formula is as follows:

[0024]

[0025] Conversion ratio R α For the original three-dimensional discrete data volume V of the physical field 3d_raw The data point position and the three-dimensional physical property field intermediate data volume V 3d_inter The mutual transformation calculation of the data point positions in the dimension direction is as follows:

[0026]

[0027] Among them, P raw_α is the original data point position of the original three-dimensional discrete data volume of the physical field, P res_α is the data point position of the intermediate data volume of the three-dimensional physical property field, R α is the transformation ratio;

[0028] Then, according to K d_α and R α The value of and the data point position calculation formula are used to calculate the value of the data point of the three-dimensional physical property field intermediate data volume between each two adjacent original data point positions of the original three-dimensional discrete data volume of the physical property field;

[0029] When K d_α When it is an integer, it means integer multiple densification, then the original data points of each original three-dimensional discrete data volume of the physical property field must be consistent with the three-dimensional physical property field intermediate data volume V 3d_interThe position of a data point in the 3D discrete data volume V is calculated by the data point position calculation formula. The value of the overlapping data point does not need to be calculated. It is directly obtained from the original physical field 3D discrete data volume V 3d_raw Copy to the 3D physical property field intermediate data volume V 3d_inter The corresponding position in

[0030] When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is greater than that of the original three-dimensional discrete data volume V of the physical field. 3d_raw When the number of data points on the boundary of the dimension is small, it means that the data points on the boundary of the dimension need to be sparse. The specific operation is as follows:

[0031] First, according to the original three-dimensional discrete data volume V 3d_raw and the 3D physical property field intermediate data volume V 3d_inter The number of data points N on the boundary of this dimension raw_α and N res_α , calculate the integer multiple sparsification judgment factor K s_α and the conversion factor R α , where K s_α The calculation formula is as follows:

[0032]

[0033] When K s_α When it is an integer, that is, (N raw_α -1) can be (N res_α -1) is divisible, indicating integer multiple sparseness, the intermediate data volume V of the three-dimensional physical property field 3d_inter The position of each data point in the original three-dimensional discrete data volume V 3d_raw If a data point in the data point overlaps, the overlap position is calculated by the data point position calculation formula, and the overlap data point value does not need to be calculated and directly copied;

[0034] When K s_α If it is not an integer, the R α Value, first use the data point position calculation formula to calculate the three-dimensional physical property field intermediate data volume V 3d_inter The original three-dimensional discrete data volume V of the physical field corresponding to the position of each data point 3d_raw Then, for the intermediate data volume V of the three-dimensional physical property field 3d_inter Any data point of the physical field is obtained through its corresponding three-dimensional discrete data volume V 3d_raw The position in the image is then searched for the two original physical field three-dimensional discrete data volumes V closest to it. 3d_raw The position of the data point in the middle is calculated by interpolation method to obtain the intermediate data volume V of the three-dimensional physical property field. 3d_inter The value of the data point.

[0035] In the above-mentioned method for arbitrary densification and sparseness processing of discrete data volume of three-dimensional physical property field, the specific steps of step 4) are as follows:

[0036] 4.1) The 3D physical property field intermediate data volume V 3d_inter The data points on the top and bottom observation surfaces are copied to the resulting three-dimensional data volume V 3d_res In the top and bottom observation surfaces;

[0037] 4.2) Based on the 3D physical property field intermediate data volume V 3d_inter And the resulting three-dimensional data volume V 3d_res The number of data points in the observation direction determines whether to perform densification or sparse processing, and calculates the integer multiple densification judgment factor K in the observation direction d_β Or the integer multiple sparseness judgment factor K in the observation direction s_β , and the observation direction change factor R β , the calculation formula is as follows:

[0038]

[0039] Where: N res_β 、N raw_β are the three-dimensional data volume V of the result 3d_res , the original three-dimensional data volume V 3d_raw The number of data points in the observation direction dimension;

[0040] The calculation formula for the data point position in the observation direction dimension is as follows:

[0041]

[0042] 4.3) According to the K calculated in step 4.2) d_β or K s_β , and the observation direction change factor R β , based on the intermediate data volume V of the three-dimensional physical property field in the observation direction 3d_inter The data of each observation surface in the calculation result is interpolated and the three-dimensional data volume V is calculated. 3d_res The data of all observation surfaces except the top and bottom observation surfaces.

[0043] In the above-mentioned arbitrary densification and sparseness processing method of the discrete data volume of the three-dimensional physical property field, the specific operation of step 4.3) is as follows:

[0044] On each data plane parallel to the observation direction, the 3D physical field intermediate data volume V 3d_inter The resulting three-dimensional data volume V 3d_res The number of columns is the same and overlaps. According to the intermediate data volume V of the three-dimensional physical property field 3d_interThe resulting three-dimensional data volume V 3d_res The one-dimensional position relationship between each column of data points in the 3D physical property field is obtained by interpolation method. 3d_inter The three-dimensional data volume V is calculated from the data point values ​​in 3d_res The values ​​of the data points in the same column;

[0045] Or on each data plane in the same direction as the observation direction, according to the intermediate data volume V of the three-dimensional physical property field 3d_inter The resulting three-dimensional data volume V 3d_res The two-dimensional positional relationship between each data point on each data surface is obtained by interpolation method from the three-dimensional physical property field intermediate data volume V 3d_inter The three-dimensional data volume V is calculated from the data point values ​​in 3d_res The value of the data point on the corresponding data surface.

[0046] Compared with the prior art, the present invention has the following beneficial effects: 1) The present invention uses a three-dimensional physical field discrete data volume of any scale as input (the minimum number of rows and columns: 2×2×2), and adopts the present invention to perform transformation processing to obtain a three-dimensional data volume output of any scale, and the number of data points in the three-dimensional directions of the input and output three-dimensional data volumes does not need to be in integer multiples (the method is compatible with multiples); 2) The present invention is an adaptive method, and no method control parameters are required when using it. The user only needs to select one-dimensional and two-dimensional interpolation calculation methods according to the physical field characteristics (the control parameters required by the interpolation method are irrelevant to this method); 3) The output results generated by the processing of the present invention It only depends on input data, and in principle no irrelevant information is added. The error only comes from the selected interpolation calculation method (the interpolation calculation method is irrelevant to the present processing method); 4) The processing calculation cost of the present invention itself is relatively small, and the interpolation calculation in the processing process is the main contributor to the calculation cost (the interpolation calculation method is irrelevant to the present processing method). The overall calculation cost is positively correlated with the scale of the transformed three-dimensional data volume, which is completely acceptable under current conditions; 5) The present invention has a framework. The one-dimensional and two-dimensional interpolation calculation methods required in the method processing process can be selected or designed according to the characteristics and requirements of the physical field characteristics and the application field itself, so that the method has high flexibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a flow chart of the present invention.

[0048] Figure 2 An example diagram of the number and location of data points in three data volumes during a complete processing process. Figure 2 (a) is the number and position diagram of data points in the original three-dimensional discrete data volume of the physical property field. Figure 2 (b) is the number and position diagram of data points in the intermediate data volume of the three-dimensional physical property field. Figure 2(c) is a graph showing the number and location of data points in the resulting three-dimensional data volume.

[0049] Figure 3 This is an example diagram of the position relationship of data points on the data surface boundary. Figure 3 (a) is an example of densification or sparseness of non-integer multiples of data points. Figure 3 (b) is an example of densification or sparseness of integer multiples of data points.

[0050] Figure 4 This is an example diagram of the position relationship of data points on the same data surface in two different data volumes. Figure 4 (a) is an example of a data surface where the data points in the row and column directions are all non-integer multiples of densification or sparseness. Figure 4 (b) is an example of a data surface in which the data points in the row direction are densified or sparse as an integer multiple, and the data points in the column direction are densified or sparse as an integer multiple.

[0051] Figure 5 This is an example diagram of data point distribution of the original three-dimensional discrete data volume sample of the physical property field in the processing example.

[0052] Figure 6 This is an example diagram of the data point distribution of the intermediate data volume of the three-dimensional physical property field in the processing example. Figure 6 (a) is an example of the intermediate data volume of the three-dimensional physical property field calculated with the Z axis as the observation direction. Figure 6 (b) is an example of the intermediate data volume of the three-dimensional physical property field calculated with the Y-axis direction as the observation direction. Figure 6 (c) is an example diagram of the intermediate data volume of the three-dimensional physical property field calculated with the X-axis direction as the observation direction.

[0053] Figure 7 An example diagram of data point distribution of a three-dimensional data volume in the processing example. DETAILED DESCRIPTION

[0054] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0055] Define the original three-dimensional discrete data volume of the physical field as V 3d_raw , the number of data points in the X, Y, and Z dimensions are N respectively raw_x 、N raw_y 、N raw_z , define the resulting three-dimensional data volume as V 3d_res , the number of data points in the X, Y, and Z dimensions are N respectively res_x 、N res_y 、N res_z .

[0056] like Figure 1 As shown, the present invention comprises the following steps:

[0057] 1) According to the application characteristics of the physical field, select an observation direction in the original three-dimensional discrete data volume of the physical field, and construct a three-dimensional physical field intermediate data volume V based on the selected observation direction. 3d_inter , three-dimensional physical property field intermediate data volume V 3d_inter In the observation direction, V 3d_raw With the same number of layers, and the intermediate data volume V of the three-dimensional physical property field 3d_inter The positions of each layer and the original three-dimensional discrete data volume V of the physical field 3d_raw The positions of each layer overlap, but the number of data points in each layer is the number of data points after arbitrary density and sparseness in the two-dimensional directions of the two-dimensional observation surface.

[0058] When the observation direction is selected as the Z direction, the observation plane is the XY plane, and the three-dimensional physical property field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. res_x 、N res_y 、N raw_z ; When the observation direction is selected as the Y direction, the observation surface is the XZ plane, and the three-dimensional physical field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. res_x 、N raw_y 、N res_z ; When the observation direction is selected as the X direction, the observation surface is the YZ plane, and the three-dimensional physical field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. raw_x 、N res_y 、N res_z .

[0059] 2) Based on the original three-dimensional discrete data volume V of the physical field 3d_raw , interpolation calculation of the three-dimensional physical property field intermediate data volume V 3d_inter The value of each discrete data point on each observation surface.

[0060] The following steps are used to calculate the three-dimensional physical property field intermediate data volume V one by one 3d_inter The value of each discrete data point on the 1st to Nth observation surfaces, where N is the total number of observation surfaces of the original three-dimensional discrete data volume of the physical property field.

[0061] 2.1) The 3D physical property field intermediate data volume V 3d_inter The i-th observation surface and the original three-dimensional discrete data volume V of the physical field 3d_raw The four boundaries of the i-th observation surface are aligned, where i∈[1,N].

[0062] 2.2) Using the interpolation method, based on the original three-dimensional discrete data volume V of the physical field 3d_rawThe boundary data of the i-th observation surface are used to calculate the intermediate data volume V of the three-dimensional physical property field. 3d_inter The data point values ​​on the four boundaries of the i-th observation surface.

[0063] Calculate the intermediate data volume V of the three-dimensional physical property field 3d_inter When the data point values ​​on the four boundaries of the i-th observation surface are calculated, the data point values ​​on the boundaries of the two dimensional directions of the observation surface are calculated respectively. The calculation method for the data point values ​​on the boundaries of each dimensional direction is as follows:

[0064] When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is the same as the original three-dimensional discrete data volume V of the physical field. 3d_raw At the same time, the original three-dimensional discrete data volume V 3d_raw The data points on the two boundaries of the dimension direction are directly copied one by one to the intermediate data volume V 3d_inter In the data points corresponding to the corresponding boundaries;

[0065] When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is greater than that of the original three-dimensional discrete data volume V of the physical field. 3d_raw When there are many data points on the boundary of this dimension, it means that the data points on the boundary of this dimension need to be densified. The specific operation is as follows:

[0066] First, according to the original three-dimensional discrete data volume V 3d_raw and the 3D physical property field intermediate data volume V 3d_inter The number of data points N on the boundary of this dimension raw_α and N res_α , calculate the integer multiple densification judgment factor K in the dimension direction d_α , the calculation formula is as follows:

[0067] Then calculate the transformation ratio R α , the calculation formula is as follows:

[0068]

[0069] Conversion ratio R α For the original three-dimensional discrete data volume V of the physical field 3d_raw The data point position and the three-dimensional physical property field intermediate data volume V 3d_inter The mutual transformation calculation of the data point positions in the dimension direction is as follows:

[0070]

[0071] Among them, P raw_αis the original data point position of the original three-dimensional discrete data volume of the physical field, P res_α is the data point position of the intermediate data volume of the three-dimensional physical property field, R α is the transformation ratio;

[0072] Then, according to K d_α and R α The value of and the data point position calculation formula are used to calculate the value of the data point of the three-dimensional physical property field intermediate data volume between each two adjacent original data point positions of the original three-dimensional discrete data volume of the physical property field;

[0073] When K d_α When it is an integer, it means integer multiple densification, then the original data points of each original three-dimensional discrete data volume of the physical property field must be consistent with the three-dimensional physical property field intermediate data volume V 3d_inter The position of a data point in the 3D discrete data volume V is calculated by the data point position calculation formula. The value of the overlapping data point does not need to be calculated. It is directly obtained from the original physical field 3D discrete data volume V 3d_raw Copy to the 3D physical property field intermediate data volume V 3d_inter The corresponding position in

[0074] When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is greater than that of the original three-dimensional discrete data volume V of the physical field. 3d_raw When the number of data points on the boundary of the dimension is small, it means that the data points on the boundary of the dimension need to be sparse. The specific operation is as follows:

[0075] First, according to the original three-dimensional discrete data volume V 3d_raw and the 3D physical property field intermediate data volume V 3d_inter The number of data points N on the boundary of this dimension raw_α and N res_α , calculate the integer multiple sparsification judgment factor K s_α and the conversion factor R α , where K s_α The calculation formula is as follows:

[0076]

[0077] When K s_α When it is an integer, that is, (N raw_α -1) can be (N res_α -1) is divisible, indicating integer multiple sparseness, the intermediate data volume V of the three-dimensional physical property field 3d_inter The position of each data point in the original three-dimensional discrete data volume V 3d_raw If a data point in the data point overlaps, the overlap position is calculated by the data point position calculation formula, and the overlap data point value does not need to be calculated and directly copied;

[0078] When K s_α If it is not an integer, the R α Value, first use the data point position calculation formula to calculate the three-dimensional physical property field intermediate data volume V 3d_inter The original three-dimensional discrete data volume V of the physical field corresponding to the position of each data point 3d_raw Then, for any three-dimensional physical property field intermediate data volume V 3d_inter The data points are obtained through the corresponding original three-dimensional discrete data volume V of the physical field 3d_raw Find the two original physical field three-dimensional discrete data volumes V closest to it. 3d_raw The position of the data point in the middle is calculated by interpolation method to obtain the intermediate data volume V of the three-dimensional physical property field. 3d_inter The value of the data point.

[0079] 2.3) Using the interpolation method, based on the original three-dimensional discrete data volume V of the physical field 3d_raw The observation surface data of the i-th observation surface and the three-dimensional physical property field intermediate data volume V 3d_inter The data point value on the boundary of the i-th observation surface is used to calculate the intermediate data volume V of the three-dimensional physical property field 3d_inter The values ​​of all non-boundary data points in the i-th observation surface.

[0080] 3) Construct a three-dimensional data volume V of the three-dimensional physical field results 3d_res , the three-dimensional data volume V of the three-dimensional physical field results 3d_res The number of layers in the observation direction is the value after arbitrary density and sparseness. The number of data points in the two dimensions of the two-dimensional observation surface of each layer is the same as the intermediate data volume V of the three-dimensional physical field. 3d_inter The number of data points in both dimensions of is the same.

[0081] 4) Based on the intermediate data volume V of the three-dimensional physical property field 3d_inter First, copy the data point values ​​of the top and bottom observation surfaces to the three-dimensional data volume V of the three-dimensional physical field result. 3d_res The three-dimensional data volume V of the interpolation calculation result is then obtained on the top and bottom observation surfaces. 3d_res The value of each discrete data point on each observation surface except the top and bottom surfaces. The specific operations are as follows:

[0082] 4.1) The 3D physical property field intermediate data volume V 3d_inter The data points on the top and bottom observation surfaces are copied to the resulting three-dimensional data volume V 3d_res In the top and bottom observation surfaces;

[0083] 4.2) Based on the 3D physical property field intermediate data volume V 3d_inter And the resulting three-dimensional data volume V 3d_resThe number of data points in the observation direction determines whether to perform densification or sparse processing, and calculates the integer multiple densification judgment factor K in the observation direction d_β Or the integer multiple sparseness judgment factor K in the observation direction s_β , and the observation direction change factor R β , the calculation formula is as follows:

[0084]

[0085] Where: N res_β 、N raw_β are the three-dimensional data volume V of the result 3d_res , the original three-dimensional data volume V 3d_raw The number of data points in the observation direction;

[0086] The calculation formula for the data point position in the observation direction dimension is as follows:

[0087]

[0088] 4.3) According to the K calculated in step 4.2) d_β or K s_β , and the observation direction change factor R β , based on the intermediate data volume V of the three-dimensional physical property field in the observation direction 3d_inter The data of each observation surface in the calculation result is interpolated and the three-dimensional data volume V is calculated. 3d_res The data of all observation surfaces except the top and bottom observation surfaces.

[0089] Step 4.3) The specific operations are as follows:

[0090] On each data plane parallel to the observation direction, the 3D physical field intermediate data volume V 3d_inter The resulting three-dimensional data volume V 3d_res The number of columns is the same and overlaps. According to the intermediate data volume V of the three-dimensional physical property field 3d_inter The resulting three-dimensional data volume V 3d_res The one-dimensional position relationship between each column of data points in the 3D physical property field is obtained by interpolation method. 3d_inter The three-dimensional data volume V is calculated from the data point values ​​in 3d_res The values ​​of the data points in the same column.

[0091] Or on each data plane in the same direction as the observation direction, according to the intermediate data volume V of the three-dimensional physical property field 3d_inter The resulting three-dimensional data volume V 3d_res The two-dimensional positional relationship between each data point on each data surface is obtained by interpolation method from the three-dimensional physical property field intermediate data volume V 3d_inter The three-dimensional data volume V is calculated from the data point values ​​in3d_res The value of the data point on the corresponding data surface.

[0092] The present invention is further described below in conjunction with a specific embodiment.

[0093] Example 1: Densification and sparseness processing of three-dimensional electromagnetic field intensity observation data

[0094] In the observation system, the number of measurement lines and measurement points is designed and selected according to the needs of the actual application. After obtaining the field strength data and calculating the data of each layer of the depth dimension, in order to obtain a better three-dimensional visualization analysis effect, it is often necessary to perform reasonable densification and sparse processing on the original three-dimensional data body, such as sparseness of the number of measurement points, densification of the number of measurement lines, densification of the number of depth layers, etc. This example takes the processing of a three-dimensional electric field intensity data body as an example to illustrate the processing process and effect of this method.

[0095] This embodiment selects an original data volume sample of size 7×101×6 (rows×columns×layers), such as Figure 5 As shown in the figure, the radius of the data point is positively correlated with the size of its data point value. In order to obtain better results in 3D visualization, it is hoped to generate a 25×51×21 (row×column×layer) result data volume based on the original data volume sample. Therefore, the original data volume sample needs to be densified in the row and layer dimensions and sparse in the column dimension.

[0096] In order to verify the arbitrariness of the method in selecting the observation direction, in this example, the three dimensions of layer, row, and column are selected as the observation direction, and three intermediate data volumes of size 25×51×6, 7×51×21, and 25×101×21 are constructed. In this example, the linear interpolation method is selected for the data point values ​​on each boundary of the intermediate observation surface, and the inverse square distance interpolation method is used for the non-boundary data point values ​​on the observation surface. The values ​​of all data points in the intermediate data volume are calculated, and the results are shown in the figure. Figure 6 As shown in (a), (b) and (c).

[0097] Construct a 25×51×21 (row×column×layer) result data volume. Based on the intermediate data volume calculated in the previous step, for simplicity, the linear interpolation method is used to calculate the values ​​of the data points on each observation surface of the result data volume. The results are as follows: Figure 7 shown.

[0098] In this example, non-integer densification is implemented in the row dimension, integer densification is implemented in the layer dimension, and integer sparsification is implemented in the column dimension. The data is processed in three observation directions, and finally a satisfactory result data volume is obtained.

Claims

1. A method for arbitrary densification and sparseness processing of a three-dimensional physical property field discrete data volume, defining the original three-dimensional physical property field discrete data volume as V 3d_raw , the number of data points in the X, Y, and Z dimensions are N respectively raw_x 、N raw_y 、N raw_z , define the resulting three-dimensional data volume as V 3d_res , the number of data points in the X, Y, and Z dimensions are N respectively res_x 、N res_y 、N res_z ; The steps include: 1) According to the application characteristics of the physical field, select an observation direction in the original three-dimensional discrete data volume of the physical field, and construct a three-dimensional physical field intermediate data volume V based on the selected observation direction. 3d_inter , three-dimensional physical property field intermediate data volume V 3d_inter In the observation direction, V 3d_raw With the same number of layers, and the 3D physical field intermediate data volume V 3d_inter The positions of each layer and the original three-dimensional discrete data volume V of the physical field 3d_raw The positions of each layer of coincide with each other, but the number of data points in each layer is the number of data points after arbitrary density and sparseness in the two-dimensional directions of the two-dimensional observation surface; 2) Based on the original three-dimensional discrete data volume V of the physical field 3d_raw , interpolation calculation of the three-dimensional physical property field intermediate data volume V 3d_inter The value of each discrete data point on each observation surface of ; 3) Construct a three-dimensional data volume V of the three-dimensional physical field results 3d_res , the three-dimensional data volume V of the three-dimensional physical field results 3d_res The number of layers in the observation direction is the value after arbitrary density and sparseness. The number of data points in the two dimensions of the two-dimensional observation surface of each layer is the same as the intermediate data volume V of the three-dimensional physical field. 3d_inter The number of data points in the two dimensions of is the same; 4) Based on the intermediate data volume V of the three-dimensional physical property field 3d_inter First, copy the data point values ​​of the top and bottom observation surfaces to the three-dimensional data volume V of the three-dimensional physical field result. 3d_res The three-dimensional data volume V of the interpolation calculation result is then obtained on the top and bottom observation surfaces. 3d_res The value of each discrete data point on each observation face except the top and bottom faces.

2. According to the method for arbitrary densification and sparseness processing of discrete data volume of three-dimensional physical property field according to claim 1, the specific operation of step 1) is as follows; When the observation direction is selected as the Z direction, the observation plane is the XY plane, and the three-dimensional physical property field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. res_x 、N res_y 、N raw_z ; When the observation direction is selected as the Y direction, the observation plane is the XZ plane, and the three-dimensional physical property field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. res_x 、N raw_y 、N res_z ; When the observation direction is selected as the X direction, the observation plane is the YZ plane, and the three-dimensional physical property field intermediate data volume V 3d_inter The number of data points in the X, Y, and Z dimensions are N respectively. raw_x 、N res_y 、N res_z .

3. According to the method for arbitrary densification and sparseness processing of discrete data volumes of three-dimensional physical property fields as described in claim 1, in step 2), the following steps are used to calculate the intermediate data volumes V of the three-dimensional physical property fields one by one: 3d_inter The value of each discrete data point on the 1st to Nth observation surfaces, where N is the total number of observation surfaces of the original three-dimensional discrete data volume of the physical property field; 2.1) The 3D physical property field intermediate data volume V 3d_inter The i-th observation surface and the original three-dimensional discrete data volume V of the physical field 3d_raw The four boundaries of the i-th observation surface are aligned, where i∈[1,N]; 2.2) Using the interpolation method, based on the original three-dimensional discrete data volume V of the physical field 3d_raw The boundary data of the i-th observation surface are used to calculate the intermediate data volume V of the three-dimensional physical property field. 3d_inter The data point values ​​on the four boundaries of the i-th observation surface; 2.3) Using the interpolation method, based on the original three-dimensional discrete data volume V of the physical field 3d_raw The observation surface data of the i-th observation surface and the three-dimensional physical property field intermediate data volume V 3d_inter The data point value on the boundary of the i-th observation surface is used to calculate the intermediate data volume V of the three-dimensional physical property field 3d_inter The values ​​of all non-boundary data points in the i-th observation surface.

4. The method for arbitrary densification and sparseness processing of discrete data volumes of three-dimensional physical property fields according to claim 3, wherein in step 2.2), the intermediate data volume V of the three-dimensional physical property field is calculated. 3d_inter When the data point values ​​on the four boundaries of the i-th observation surface are calculated, the data point values ​​on the boundaries of the two dimensional directions of the observation surface are calculated respectively. The calculation method for the data point values ​​on the boundaries of each dimensional direction is as follows: When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is the same as the original three-dimensional discrete data volume V of the physical field. 3d_raw At the same time, the original three-dimensional discrete data volume V 3d_raw The data points on the two boundaries of the dimension direction are directly copied one by one to the intermediate data volume V 3d_inter In the data points corresponding to the corresponding boundaries; When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is greater than that of the original three-dimensional discrete data volume V of the physical field. 3d_raw When there are many data points on the boundary of this dimension, it means that the data points on the boundary of this dimension need to be densified. The specific operation is as follows: First, according to the original three-dimensional discrete data volume V 3d_raw and the 3D physical property field intermediate data volume V 3d_inter The number of data points N on the boundary of this dimension raw_α and N res_α , calculate the integer multiple densification judgment factor K in the dimension direction d_α , the calculation formula is as follows: Then calculate the transformation ratio R α , the calculation formula is as follows: Conversion ratio R α For the original three-dimensional discrete data volume V of the physical field 3d_raw The data point position and the three-dimensional physical property field intermediate data volume V 3d_inter The mutual transformation calculation of the data point positions in the dimension direction is as follows: in, P raw_α is the original data point position of the original three-dimensional discrete data volume of the physical field, P res_α is the data point position of the intermediate data volume of the three-dimensional physical property field, R α is the transformation ratio; Then, according to K d_α and R α The value of and the data point position calculation formula are used to calculate the value of the data point of the three-dimensional physical property field intermediate data volume between each two adjacent original data point positions of the original three-dimensional discrete data volume of the physical property field; When K d_α When it is an integer, it means integer multiple densification, then the original data points of each original three-dimensional discrete data volume of the physical property field must be consistent with the three-dimensional physical property field intermediate data volume V 3d_inter The position of a data point in the 3D discrete data volume V is calculated by the data point position calculation formula. The value of the overlapping data point does not need to be calculated. It is directly obtained from the original physical field 3D discrete data volume V 3d_raw Copy to the 3D physical property field intermediate data volume V 3d_inter The corresponding position in When the intermediate data volume V 3d_inter The number of data points on the boundary of this dimension is greater than that of the original three-dimensional discrete data volume V of the physical field. 3d_raw When the number of data points on the boundary of the dimension is small, it means that the data points on the boundary of the dimension need to be sparse. The specific operation is as follows: First, according to the original three-dimensional discrete data volume V 3d_raw and the 3D physical property field intermediate data volume V 3d_inter The number of data points N on the boundary of this dimension raw_α and N res_α , calculate the integer multiple sparse judgment factor K s_α and the conversion factor R α , where K s_α The calculation formula is as follows: When K s_α When it is an integer, that is, (N raw_α -1) can be (N res_α -1) is divisible, indicating integer multiple sparseness, the intermediate data volume V of the three-dimensional physical property field 3d_inter The position of each data point in the original three-dimensional discrete data volume V 3d_raw If a data point in the data point overlaps, the overlap position is calculated by the data point position calculation formula, and the overlap data point value does not need to be calculated and directly copied; When K s_α If it is not an integer, the R α Value, first use the data point position calculation formula to calculate the three-dimensional physical property field intermediate data volume V 3d_inter The original three-dimensional discrete data volume V of the physical field corresponding to the position of each data point 3d_raw Then, for the intermediate data volume V of the three-dimensional physical property field 3d_inter Any data point of the physical field is obtained through its corresponding three-dimensional discrete data volume V 3d_raw Find the two original physical field three-dimensional discrete data volumes V closest to it. 3d_raw The position of the data point in the middle is calculated by interpolation method to obtain the intermediate data volume V of the three-dimensional physical property field. 3d_inter The value of the data point.

5. According to the method for arbitrary densification and sparseness processing of discrete data volume of three-dimensional physical property field of claim 4, the specific steps of step 4) are as follows: 4.1) The 3D physical property field intermediate data volume V 3d_inter The data points on the top and bottom observation surfaces are copied to the resulting three-dimensional data volume V 3d_res In the top and bottom observation surfaces; 4.2) Based on the 3D physical property field intermediate data volume V 3d_inter And the resulting three-dimensional data volume V 3d_res The number of data points in the observation direction determines whether to perform densification or sparse processing, and calculates the integer multiple densification judgment factor K in the observation direction d_β Or the integer multiple sparseness judgment factor K in the observation direction s_β , and the observation direction change factor R β , the calculation formula is as follows: in: N res_β 、N raw_β are the three-dimensional data volume V of the result 3d_res , the original three-dimensional data volume V 3d_raw The number of data points in the observation direction dimension; The calculation formula for the data point position in the observation direction dimension is as follows: 4.3) According to the K calculated in step 4.2) d_β or K s_β , and the observation direction change factor R β , based on the intermediate data volume V of the three-dimensional physical property field in the observation direction 3d_inter The data of each observation surface in the calculation result is interpolated and the three-dimensional data volume V is calculated. 3d_res The data of all observation surfaces except the top and bottom observation surfaces.

6. According to the method for arbitrary densification and sparseness processing of discrete data volume of three-dimensional physical property field according to claim 5, the specific operation of step 4.3) is as follows: On each data plane parallel to the observation direction, the intermediate data volume V of the three-dimensional physical property field 3d_inter The resulting three-dimensional data volume V 3d_res The number of columns is the same and overlaps. According to the intermediate data volume V of the three-dimensional physical property field 3d_inter The resulting three-dimensional data volume V 3d_res The one-dimensional position relationship between each column of data points in the 3D physical property field is obtained by interpolation method. 3d_inter The three-dimensional data volume V is calculated from the data point values ​​in 3d_res The values ​​of the data points in the same column; Or on each data plane in the same direction as the observation direction, according to the intermediate data volume V of the three-dimensional physical property field 3d_inter The resulting three-dimensional data volume V 3d_res The two-dimensional positional relationship between each data point on each data surface is obtained by interpolation method from the three-dimensional physical property field intermediate data volume V 3d_inter The three-dimensional data volume V is calculated from the data point values ​​in 3d_res The value of the data point on the corresponding data surface.

Images