// David Eberly, Geometric Tools, Redmond WA 98052 // Copyright (c) 1998-2020 // Distributed under the Boost Software License, Version 1.0. // https://www.boost.org/LICENSE_1_0.txt // https://www.geometrictools.com/License/Boost/LICENSE_1_0.txt // Version: 4.0.2019.08.13 #pragma once #include #include #include #include namespace WwiseGTE { // The quadratic fit is // 0 = C[0] + C[1]*X + C[2]*Y + C[3]*Z + C[4]*X^2 + C[5]*Y^2 // + C[6]*Z^2 + C[7]*X*Y + C[8]*X*Z + C[9]*Y*Z // subject to Length(C) = 1. Minimize E(C) = C^t M C with Length(C) = 1 // and M = (sum_i V_i)(sum_i V_i)^t where // V = (1, X, Y, Z, X^2, Y^2, Z^2, X*Y, X*Z, Y*Z) // The minimum value is the smallest eigenvalue of M and C is a // corresponding unit length eigenvector. // // Input: // n = number of points to fit // p[0..n-1] = array of points to fit // // Output: // c[0..9] = coefficients of quadratic fit (the eigenvector) // return value of function is nonnegative and a measure of the fit // (the minimum eigenvalue; 0 = exact fit, positive otherwise) // // Canonical forms. The quadratic equation can be factored into // P^T A P + B^T P + K = 0 where P = (X,Y,Z), K = C[0], // B = (C[1],C[2],C[3]), and A is a 3x3 symmetric matrix with // A00 = C[4], A11 = C[5], A22 = C[6], A01 = C[7]/2, A02 = C[8]/2, and // A12 = C[9]/2. Matrix A = R^T D R where R is orthogonal and D is // diagonal (using an eigendecomposition). Define V = R P = (v0,v1,v2), // E = R B = (e0,e1,e2), D = diag(d0,d1,d2), and f = K to obtain // d0 v0^2 + d1 v1^2 + d2 v^2 + e0 v0 + e1 v1 + e2 v2 + f = 0 // The characterization depends on the signs of the d_i. template class ApprQuadratic3 { public: Real operator()(int numPoints, Vector3 const* points, Real coefficients[10]) { Matrix<10, 10, Real> A; // constructor sets A to zero for (int i = 0; i < numPoints; ++i) { Real x = points[i][0]; Real y = points[i][1]; Real z = points[i][2]; Real x2 = x * x; Real y2 = y * y; Real z2 = z * z; Real xy = x * y; Real xz = x * z; Real yz = y * z; Real x3 = x * x2; Real xy2 = x * y2; Real xz2 = x * z2; Real x2y = x * xy; Real x2z = x * xz; Real xyz = x * y * z; Real y3 = y * y2; Real yz2 = y * z2; Real y2z = y * yz; Real z3 = z * z2; Real x4 = x * x3; Real x2y2 = x * xy2; Real x2z2 = x * xz2; Real x3y = x * x2y; Real x3z = x * x2z; Real x2yz = x * xyz; Real y4 = y * y3; Real y2z2 = y * yz2; Real xy3 = x * y3; Real xy2z = x * y2z; Real y3z = y * y2z; Real z4 = z * z3; Real xyz2 = x * yz2; Real xz3 = x * z3; Real yz3 = y * z3; A(0, 1) += x; A(0, 2) += y; A(0, 3) += z; A(0, 4) += x2; A(0, 5) += y2; A(0, 6) += z2; A(0, 7) += xy; A(0, 8) += xz; A(0, 9) += yz; A(1, 4) += x3; A(1, 5) += xy2; A(1, 6) += xz2; A(1, 7) += x2y; A(1, 8) += x2z; A(1, 9) += xyz; A(2, 5) += y3; A(2, 6) += yz2; A(2, 9) += y2z; A(3, 6) += z3; A(4, 4) += x4; A(4, 5) += x2y2; A(4, 6) += x2z2; A(4, 7) += x3y; A(4, 8) += x3z; A(4, 9) += x2yz; A(5, 5) += y4; A(5, 6) += y2z2; A(5, 7) += xy3; A(5, 8) += xy2z; A(5, 9) += y3z; A(6, 6) += z4; A(6, 7) += xyz2; A(6, 8) += xz3; A(6, 9) += yz3; A(9, 9) += y2z2; } A(0, 0) = static_cast(numPoints); A(1, 1) = A(0, 4); A(1, 2) = A(0, 7); A(1, 3) = A(0, 8); A(2, 2) = A(0, 5); A(2, 3) = A(0, 9); A(2, 4) = A(1, 7); A(2, 7) = A(1, 5); A(2, 8) = A(1, 9); A(3, 3) = A(0, 6); A(3, 4) = A(1, 8); A(3, 5) = A(2, 9); A(3, 7) = A(1, 9); A(3, 8) = A(1, 6); A(3, 9) = A(2, 6); A(7, 7) = A(4, 5); A(7, 8) = A(4, 9); A(7, 9) = A(5, 8); A(8, 8) = A(4, 6); A(8, 9) = A(6, 7); A(9, 9) = A(5, 6); for (int row = 0; row < 10; ++row) { for (int col = 0; col < row; ++col) { A(row, col) = A(col, row); } } Real invNumPoints = (Real)1 / static_cast(numPoints); for (int row = 0; row < 10; ++row) { for (int col = 0; col < 10; ++col) { A(row, col) *= invNumPoints; } } SymmetricEigensolver es(10, 1024); es.Solve(&A[0], +1); es.GetEigenvector(0, &coefficients[0]); // For an exact fit, numeric round-off errors might make the // minimum eigenvalue just slightly negative. Return the absolute // value because the application might rely on the return value // being nonnegative. return std::fabs(es.GetEigenvalue(0)); } }; // If you think your points are nearly spherical, use this. The sphere is // of form C'[0]+C'[1]*X+C'[2]*Y+C'[3]*Z+C'[4]*(X^2+Y^2+Z^2) where // Length(C') = 1. The function returns // C = (C'[0]/C'[4],C'[1]/C'[4],C'[2]/C'[4],C'[3]/C'[4]), so the fitted // sphere is C[0]+C[1]*X+C[2]*Y+C[3]*Z+X^2+Y^2+Z^2. The center is // (xc,yc,zc) = -0.5*(C[1],C[2],C[3]) and the radius is // r = sqrt(xc*xc+yc*yc+zc*zc-C[0]). template class ApprQuadraticSphere3 { public: Real operator()(int numPoints, Vector3 const* points, Sphere3& sphere) { Matrix<5, 5, Real> A; // constructor sets A to zero for (int i = 0; i < numPoints; ++i) { Real x = points[i][0]; Real y = points[i][1]; Real z = points[i][2]; Real x2 = x * x; Real y2 = y * y; Real z2 = z * z; Real xy = x * y; Real xz = x * z; Real yz = y * z; Real r2 = x2 + y2 + z2; Real xr2 = x * r2; Real yr2 = y * r2; Real zr2 = z * r2; Real r4 = r2 * r2; A(0, 1) += x; A(0, 2) += y; A(0, 3) += z; A(0, 4) += r2; A(1, 1) += x2; A(1, 2) += xy; A(1, 3) += xz; A(1, 4) += xr2; A(2, 2) += y2; A(2, 3) += yz; A(2, 4) += yr2; A(3, 3) += z2; A(3, 4) += zr2; A(4, 4) += r4; } A(0, 0) = static_cast(numPoints); for (int row = 0; row < 5; ++row) { for (int col = 0; col < row; ++col) { A(row, col) = A(col, row); } } Real invNumPoints = (Real)1 / static_cast(numPoints); for (int row = 0; row < 5; ++row) { for (int col = 0; col < 5; ++col) { A(row, col) *= invNumPoints; } } SymmetricEigensolver es(5, 1024); es.Solve(&A[0], +1); Vector<5, Real> evector; es.GetEigenvector(0, &evector[0]); // TODO: Guard against zero divide? Real inv = (Real)1 / evector[4]; Real coefficients[4]; for (int row = 0; row < 4; ++row) { coefficients[row] = inv * evector[row]; } sphere.center[0] = (Real)-0.5 * coefficients[1]; sphere.center[1] = (Real)-0.5 * coefficients[2]; sphere.center[2] = (Real)-0.5 * coefficients[3]; sphere.radius = std::sqrt(std::fabs(Dot(sphere.center, sphere.center) - coefficients[0])); // For an exact fit, numeric round-off errors might make the // minimum eigenvalue just slightly negative. Return the // absolute value because the application might rely on the // return value being nonnegative. return std::fabs(es.GetEigenvalue(0)); } }; }