CN110321619B - A Parametric Custom Model Generation Method Based on Sound Data - Google Patents

Abstract

The present invention relates to the technical field of model generation methods, and discloses a parameterized custom model generation method based on sound data for generating a sound model, comprising the following steps: 1) determining a plane U, and constructing four overlapping planes on the plane U Circle A, circle B, circle C and circle D; 2), divide circle A, circle B, circle C and circle D at equal distances respectively, circle A is set with a dividing point, and circle B is set with b dividing points , circle C is set with c dividing points, and circle D is set with d dividing points; by adjusting the values of a, b, c, and d, the adjustment and switching of the curve function f(x) and the curve function g(x), By changing the values of non-fixed parameters, this allows the use of parametric solutions to make private custom design more efficient, and the algorithm and conversion process from input sound data to the final model breaks through the high time-consuming manual design modeling in the past, Constraints such as design thinking limitations make precise customization possible for a large number of people at the same time.

Description

A Parametric Custom Model Generation Method Based on Sound Data

technical field

The patent of the present invention relates to the technical field of model generation methods, in particular, to a parameterized customized model generation method based on sound data.

Background technique

Model making based on sound data refers to taking the PCM value of the recorded sound and manually processing it, and then generating the corresponding sound model.

At present, the model generation based on sound data requires people to extract the key information in the sound data, and then extract the sound data through the key information in the sound data, and then generate a corresponding sound data after a long process of processing. sound model.

In the prior art, the model generation based on sound data is still not customizable, and only a corresponding single model can be generated based on the designed model of the existing parameters; Time-consuming, and the resulting models are sub-optimal; design thinking affects the final product.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method for generating a parametric custom model based on sound data, aiming to solve the problem that the model based on sound data in the prior art is not customizable.

The present invention is realized like this, and the parametric custom model generation method based on sound data is used for generating sound model, comprising the following steps:

1), determine plane U, and construct four overlapping circles A, circle B, circle C and circle D on described plane U;

2), the circle A, the circle B, the circle C, and the circle D are respectively equidistantly divided, the circle A is provided with a dividing points, and the circle B is provided with b dividing points , the circle C is provided with c dividing points, and the circle D is provided with d dividing points;

3), extract the peak value and the frequency from the data extracted by the PCM value of the sound, set a, b, c and d dividing points respectively from the frequency range of the PCM value of the sound, thereby obtaining a, b, and d in the frequency range. The frequency values corresponding to the c and d division points, and the peak value corresponding to the frequency can be obtained from the original PCM value; input the peak values and frequencies of the a, b, c and d division points of the PCM value of the sound to the The smooth waveforms a1, b1, c1 and d1 are obtained by calculation in the respective curve functions f(X), so as to obtain the vertical rise data corresponding to frequencies other than the dividing point on the smooth waveform;

4) Substitute the frequencies in the PCM value of the sound into the curve functions f _a1 (x), f corresponding to the gentle waveform a1, the smooth waveform a2, the smooth waveform a3, and the smooth waveform a4, respectively In _b1 (x), f _c1 (x) and f _d1 (x), it can be obtained that a division point is in the flat waveform a1, b division points are in the flat waveform b1, and c division points are in the flat waveform The points of the waveform c1 and the d division points in the intercepted interval of the gentle waveform d1 are respectively in the vertical rise data relative to the Z axis, and are substituted into circle A, circle B, circle C and circle D respectively. , so that circle A, circle B, circle C and circle D can move up on the plane U correspondingly, and then use the line to close the series connection, and the circular waveform curves A01, B01, C01 and D01 can be obtained;

5) Scale the curve A01 according to the scale value a01, scale the curve B01 according to the scale value b01, scale the curve C01 according to the scale value c01, scale the curve D01 according to the scale value d01, The relationship between the proportional values is: b01>a01>d01>c01;

6), after the curve A01, the curve B01, the curve C01 and the curve D01 are rearranged according to the stacking order of the curve B01, the curve A01, the curve D01 and the curve C01 to generate the curved surface S;

7), the curved surface S is uniformly cut in the UV direction, and the number of cuts in the two directions is u and v, and the u cutting surfaces and the v cutting surfaces are staggered to form divided sub-block surfaces, and the curved surface is After S is divided, a new curved surface S01 is generated, and a dividing line is formed in the U direction;

8), adjust the interval relationship of the dividing line by the curve function g(x), and split the interval line of the curved surface S01 after the division, and the number is u, and the interval line of the curved surface S01 includes the described surface of the number u. The front boundary line SLF and the back margin line SLB corresponding to the sub-block surface;

9), on the front boundary line SLF and the back margin line SLB, the number of equidistant generation on the line is p points respectively, and the p points on the front boundary line SLF are respectively defined as SLF_P1, SLF_P2, SLF_P3 , ..., SLF_Pp, the p points on the back margin line SLB are respectively defined as SLB_P1, SLB_P2, SLB_P3, ..., SLB_Pp;

10), connecting each SLF_P point and each SLB_P point to generate a straight line SLFB_PL whose quantity is p*u, and taking the midpoint of each straight line SLFB_PL as SLFB_PL_P;

11), calculate the distance SLFB_PL_PA_d from each point SLFB_PL_P to curve A respectively, set the distance that each SLFB_PL_P point drops on the Z axis according to the value of this distance, and set the dropped point as SLFB_PL_PD;

12), connect SLFB_PL_PD with adjacent SLF_P and SLB_P to form a curve SL_C, and the connection order is SLF_P, SLFB_PL_PD, SLB_P;

13), the described SL_C is grouped according to the mode of UV segmentation, is divided into the group that quantity is u, and the described SL_C in each group is connected to form surface SL_S, and all described SL_S are combined to form a function model;

Further, the a dividing points, the b dividing points, the c dividing points and the d dividing points in step 2) are set to be equal to each other.

Further, the curve function f _(X) described in step 3) is set to sin(ax+b), where x is the frequency, a and b are variable values in the function, when entering x and the corresponding f(x) ) value is obtained.

其中u为步骤8)中描述的拆分数量，p_i为相对间隔距 离，g(x)为距离值总和，x是相对圆心的距离。Further, the curve function g(x) described in step 8) is set as a Bayesian curve equation, and the equation is specifically

where u is the number of splits described in step 8), pi is the relative separation distance, _g (x) is the sum of the distance values, and x is the distance relative to the center of the circle.

Further, a and b in the curve functions corresponding to the b dividing points, the c dividing points and the d dividing points respectively are set to 0.

Further, the vertical rising data corresponding to the b dividing points, the c dividing points, and the d dividing points respectively are 0.

Further, each of the distances SLFB_PL_PA_d is equal to the distance that each SLFB_PL_P point falls on the Z axis.

Further, the front boundary line SLF and the back margin line SLB are spaced coincident with each other.

Further, the p dividing points on the front boundary line SLF and the p dividing points on the back margin line SLB overlap at intervals.

Further, the waveform curve on the circle A is provided with head and tail points, and the vertical rise data before and after the head and tail points are the f(x) of the first point at1 and the tail point at2 of the waveform a1 obtained in step 3). )value.

Compared with the prior art, the method for generating a parameterized customized model based on sound data provided by the present invention, by adjusting the values of a, b, c, and d, and adjusting the curve function f(x) and the curve function g(x) By changing the value of non-fixed parameters, this allows the use of parametric solutions to make private custom design more efficient, and the algorithm and conversion process from input sound data to the final model to break through the manual design and construction of the past. The high time-consuming of modeling and the limitations of design thinking make it possible to precisely customize for a large number of people at the same time.

Description of drawings

Fig. 1 is a flow chart of a method for generating a parametric custom model based on sound data provided by the present invention.

Detailed ways

In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

The implementation of the present invention will be described in detail below with reference to specific embodiments.

The same or similar numbers in the drawings of this embodiment correspond to the same or similar components; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", etc. The indicated orientation or positional relationship is based on the orientation or positional relationship shown in the accompanying drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the indicated device or element must have a specific orientation or a specific orientation. structure and operation, so the terms describing the positional relationship in the drawings are only used for exemplary illustration and should not be construed as a limitation on the present patent. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific situations.

Referring to FIG. 1, it is a preferred embodiment provided by the present invention.

A parametric custom model generation method based on sound data is used to generate a sound model, including the following steps:

1), determine plane U, and construct four overlapping circles A, circle B, circle C and circle D on plane U;

2), perform equidistant segmentation on circle A, circle B, circle C and circle D respectively, circle A is set with a split point, circle B is set with b split points, circle C is set with c split points, circle D There are d split points;

3), extract the peak value and the frequency from the data extracted by the PCM value of the sound, and set a, b, c and d division points respectively from the frequency range of the PCM value of the sound, so as to obtain a, b, c and d in the frequency range. The frequency values corresponding to the d division points, and the peak value corresponding to the frequency can be obtained from the original PCM value; input the peak values and frequencies of the a, b, c and d division points of the PCM value of the sound into the respective curve functions In f(X), the smooth waveforms a1, b1, c1 and d1 are obtained by calculation, so as to obtain the vertical rise data corresponding to frequencies other than the dividing point on the smooth waveform;

4) Substitute the frequencies in the PCM value of the sound into the curve functions f _a1 (x), f _b1 (x), f _c1 (x) corresponding to the flat waveform a1, the flat waveform a2, the flat waveform a3 and the flat waveform a4 respectively. And in f _d1 (x), it can be obtained that a division point is in the flat waveform a1, b division points are in the flat waveform b1, c division points are in the flat waveform c1, and d division points are intercepted in the flat waveform d1. The points of the interval are respectively in the vertical rise data relative to the Z axis, and they are substituted into circle A, circle B, circle C and circle D respectively, so that circle A, circle B, circle C and circle D can correspond on the plane U Move up, and then use lines to close and connect in series, you can get circular waveform curves A01, B01, C01 and D01;

5) Scale the curve A01 according to the scale value a01, scale the curve B01 according to the scale value b01, scale the curve C01 according to the scale value c01, scale the curve D01 according to the scale value d01, the relationship between the scale values is: b01>a01>d01>c01;

6), after the curve A01, the curve B01, the curve C01 and the curve D01 are rearranged according to the stacking order of the curve B01, the curve A01, the curve D01 and the curve C01 to generate the curved surface S;

7), the surface S is uniformly cut in the UV direction, the number of cuts in the two directions is u and v, the u cutting surfaces and the v cutting surfaces are staggered to form divided sub-block surfaces, and the curved surface S is divided to generate a new sub-block surface. The curved surface S01 is formed with a dividing line in the U direction;

8), adjust the spacing relationship of the dividing lines through the curve function g(x), and split the spacing lines of the divided curved surface S01, the number is u, and the spacing lines of the curved surface S01 include the corresponding sub-block surfaces of the number u. Front border line SLF and back margin line SLB;

9) On the front boundary line SLF and the back margin line SLB, the number of p points is evenly generated on the line at equal distances, and the p points on the front boundary line SLF are respectively defined as SLF_P1, SLF_P2, SLF_P3, ..., SLF_Pp , the p points on the back margin line SLB are respectively defined as SLB_P1, SLB_P2, SLB_P3, ..., SLB_Pp;

10), connect each SLF_P point and each SLB_P point to generate a straight line SLFB_PL whose quantity is p*u, and take the midpoint of each straight line SLFB_PL as SLFB_PL_P;

11), calculate the distance SLFB_PL_PA_d from each point SLFB_PL_P to curve A respectively, set the distance that each SLFB_PL_P point drops on the Z axis according to the value of this distance, and set the dropped point as SLFB_PL_PD;

12), connect SLFB_PL_PD with adjacent SLF_P and SLB_P to form a curve SL_C, and the connection order is SLF_P, SLFB_PL_PD, SLB_P;

13), SL_C is grouped according to the mode of UV segmentation, divided into groups of u, and the SL_C in each group is connected to form surface SL_S, and all SL_S are combined to form a function model;

By adjusting the values of a, b, c, d, and switching between different adjustments to the curve function f(x) and the curve function g(x), by changing the values of the non-fixed parameters, this allows the use of parametric solutions It can make private custom design more efficient. The algorithm and conversion process from the input sound data to the final model can break through the constraints of high time-consuming manual design and modeling and the limitations of design thinking in the past, making it possible to accurately customize for a large number of people at the same time. .

Specifically, the a division points, b division points, c division points and d division points in step 2) are set to be equal to each other; preferably, a can be set to 70, similarly, b, c, d are respectively It can be set to 70, 70, 70, so that the subsequent model generation can be more beautiful and more accurate, and there will be no problem of dislocation generation.

Specifically, in step 3), the curve function f _(X) is preferably set as sin(ax+b), where x is the frequency, a and b are variable values in the function, and when x and the corresponding f(x) are entered value; and setting to sin(ax+b) is only a preferred result of the curve function f ( _X ); the curve function f _(X) can be of the form sin(ax+b)+cos(wx+v) Any trigonometric function of ; by setting the trigonometric function sin(ax+b), the waveform a1 can always be flat without the problem of fault.

In addition, the curve function f _(X) can also be a Fourier series; when it is set to a Fourier series, the waveform a1 can always be flat and no fault occurs.

其中u为步骤8)中描述的拆分数量，p_i为相对间隔距离， g(x)为距离值总和，x是相对圆心的距离；通过设置贝叶斯曲线方程，可以轻松 地调节分割线的间隔关系，而后可以进一步地获取数量u的子块面。Specifically, in step 8), the curve function g(x) is set as a Bayesian curve equation, and the equation is specifically

where u is the number of splits described in step 8), pi is the relative separation distance, _g (x) is the sum of the distance values, and x is the distance from the center of the circle; the splitting line can be easily adjusted by setting the Bayesian curve equation The interval relationship of , and then the sub-block faces of the number u can be further obtained.

In addition, setting the curve function g(x) as a Bayesian curve equation is also a preferred solution, and can also be set as a hyperbolic equation or an ellipse equation; a variety of optional curve equations can be further customized for further improvement.

Specifically, a and b in the curve functions corresponding to the b dividing points, the c dividing points and the d dividing points respectively are set to 0; and setting the corresponding a and b to 0 is a preferred solution, rather than a Necessary settings in order to generate custom models that are more specific and beautiful.

Specifically, the vertical rising data corresponding to the b dividing points, the c dividing points, and the d dividing points respectively are 0; similarly, since both a and b are set to 0, the vertical rising data is also obtained because of the The curve function does not change, so it is also 0.

Specifically, each distance SLFB_PL_PA_d is equal to the distance that each SLFB_PL_P point descends on the Z axis; by decreasing the distance of SLFB_PL_PA_d, a single sub-block face is recessed, which is also a preferred solution here, and it is also possible to choose not to recess.

Specifically, the front boundary line SLF and the back margin line SLB overlap at intervals; the previous front boundary line and the previous back margin line are coincident; by setting the two as the same body, it is ensured that the points generated by the subsequent segmentation can be coincident. .

Specifically, the p dividing points on the front boundary line SLF and the p dividing points on the back margin line SLB overlap at intervals; by setting the two to overlap, it is ensured that the subsequent connection line can be a straight line, and the two adjacent SLFB_PL and phase The connected front boundary line SLF and the back margin line SLB will not form a distorted slope. In addition to ensuring the beauty of the generated model, it also ensures the purpose of customization; but the p dividing points on the front boundary line SLF and the back It is also a preferred solution to overlap the p split points on the distance line SLB, so that it can further improve the degree of customization

Specifically, the waveform curve on the circle A is provided with head and tail points, and the vertical rise data before and after the head and tail points are the f(x) values of the first point at1 and the tail point at2 of the waveform a1 obtained in step 3); by determining the first point and The f(x) value of the tail point ensures the ability to customize.

Specifically, parametric domain variables of parametric surfaces are generally expressed by UV letters, such as parametric surfaces F(u, v). Therefore, the commonly called three-dimensional surface is essentially two-dimensional, and the space in which it is embedded is three-dimensional. All surfaces that can be expressed by F(u,v) are parametric surfaces, such as NURBS surfaces. For a triangular mesh, if it can be mapped one-to-one with the parameter plane, then it is also parameterized. This mapping is UV unwrapping.

The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention shall be included in the protection of the present invention. within the range.

Claims (10)

1. a parametric custom model generation method based on sound data, for generating a sound model, is characterized in that, comprises the following steps: 1), determine plane U, and construct four overlapping circles A, circle B, circle C and circle D on said plane U; 2), the circle A, the circle B, the circle C and the circle D are respectively equidistantly divided, the circle A is provided with a dividing points, and the circle B is provided with b dividing points , the circle C is provided with c dividing points, and the circle D is provided with d dividing points; 3), extract the peak value and the frequency from the data extracted by the PCM value of the sound, and set a, b, c and d dividing points respectively from the frequency range of the PCM value of the sound, thereby obtaining a, b, and d in the frequency range. The frequency values corresponding to the c and d division points, and the peak value corresponding to the frequency can be obtained from the original PCM value; input the peak values and frequencies of the a, b, c and d division points of the PCM value of the sound to the The smooth waveforms a1, b1, c1 and d1 are obtained by calculation in the respective curve functions f(X), so as to obtain the vertical rise data corresponding to frequencies other than the dividing point on the smooth waveform; 4) Substitute the frequencies in the PCM value of the sound into the curve functions f _a1 (x), f corresponding to the gentle waveform a1, the smooth waveform a2, the smooth waveform a3, and the smooth waveform a4, respectively In _b1 (x), f _c1 (x) and f _d1 (x), it can be obtained that a division points are in the smooth waveform a1, b division points are in the smooth waveform b1, and c division points are in the smooth waveform The points of the waveform c1 and the d division points in the intercepted interval of the gentle waveform d1 are respectively in the vertical rise data relative to the Z axis, and are substituted into circle A, circle B, circle C and circle D respectively. , so that circle A, circle B, circle C and circle D can move up on plane U correspondingly, and then use lines to close and connect in series, and circular waveform curves A01, B01, C01 and D01 can be obtained; 5) Scale the curve A01 according to the scale value a01, scale the curve B01 according to the scale value b01, scale the curve C01 according to the scale value c01, scale the curve D01 according to the scale value d01, The relationship between the proportional values is: b01>a01>d01>c01; 6), after the curve A01, the curve B01, the curve C01 and the curve D01 are rearranged according to the stacking order of the curve B01, the curve A01, the curve D01 and the curve C01 to generate the curved surface S; 7), the curved surface S is uniformly cut in the UV direction, the number of cuts in the two directions is u and v, the u cutting surfaces and the v cutting surfaces are staggered to form divided sub-block surfaces, and the curved surface S is divided to generate The new curved surface S01 is formed with a dividing line in the U direction; 8), adjust the interval relationship of the dividing line by the curve function g(x), and split the interval line of the curved surface S01 after the division, and the number is u, and the interval line of the curved surface S01 includes the described surface of the number u. The front boundary line SLF and the back margin line SLB corresponding to the sub-block surface; 9), on the front boundary line SLF and the back margin line SLB, the number of equidistant generation on the line is respectively p points, and the p points on the front boundary line SLF are respectively defined as SLF_P1, SLF_P2, SLF_P3 , ..., SLF_Pp, the p points on the back margin line SLB are respectively defined as SLB_P1, SLB_P2, SLB_P3, ..., SLB_Pp; 10), connect each SLF_P point and each SLB_P point to generate a straight line SLFB_PL whose quantity is p*u, and take the midpoint of each straight line SLFB_PL as SLFB_PL_P; 11) Calculate the distance SLFB_PL_PA_d from each point SLFB_PL_P to curve A, set the distance that each SLFB_PL_P point falls on the Z axis according to the value of this distance, and set the dropped point as SLFB_PL_PD; 12), connect SLFB_PL_PD with adjacent SLF_P, SLB_P to form a curve SL_C, the connection order is SLF_P, SLFB_PL_PD, SLB_P; 13) Group the SL_Cs according to UV segmentation, and divide them into groups of u, and connect the SL_Cs in each group to form a curved surface SL_S, and combine all the SL_S to form a function model. 2. The method for generating a parameterized custom model based on sound data as claimed in claim 1, wherein in step 2) the a dividing points, the b dividing points, the c dividing points and the The d division points are set equal to each other. 3. the parametric custom model generation method based on sound data as claimed in claim 1, is characterized in that, described curve function f (X) in step 3) is set to sin (ax+b), wherein x is frequency, a and b are variable values in the function, and they are obtained by entering x and the corresponding f(x) value. 4. the parametric custom model generation method based on sound data as claimed in claim 1, it is characterised in that the curve function g (x) described in step 8) is set to a Bayesian curve equation, and the equation is specifically

where u is the number of splits described in step 8), pi is the relative separation distance, _g (x) is the sum of the distance values, and x is the distance relative to the center of the circle. 5. The method for generating a parametric custom model based on sound data according to any one of claims 1 to 3, wherein the b split points, the c split points, and the d split points are respectively Both a and b in the corresponding curve function are set to 0. 6. The method for generating a parameterized custom model based on sound data according to claim 5, wherein the vertical division points corresponding to the b dividing points, the c dividing points and the d dividing points respectively Upward data is 0. 7 . The method for generating a parametric custom model based on sound data according to claim 6 , wherein each of the distances SLFB_PL_PA_d is equal to the distance that each SLFB_PL_P point falls on the Z axis. 8 . 8. The method for generating a parametric custom model based on sound data according to any one of claims 1-3, wherein the front boundary line SLF and the rear margin line SLB overlap at intervals. 9. The method for generating a parametric custom model based on sound data according to claim 8, wherein the p dividing points on the front boundary line SLF and the p dividing points on the back margin line SLB are spaced apart coincide. 10. The method for generating a parametric custom model based on sound data according to any one of claims 1 to 3, wherein the waveform curve on the circle A is provided with head and tail points, and the front and rear of the head and tail points are described The vertical rise data is the f(x) value of the first point at1 and the tail point at2 of the waveform a1 obtained in step 3).

Images