Method and apparatus for interactive curve generation

Abstract

A system of curve generation takes a sequence of control points and constraint codes for each control point, and outputs a curve in which each of the constraints is satisfied. The set of constraints is chosen from the tangent angle, curvature, first derivative of curvature, and second derivative of curvature. The interactive curve design uses as its primitive, a curve whose curvature is a polynomial function of arclength (whose intrinsic equation is a polynomial). At each control point, a choice of G2 curvature continuity (tangent angle and curvature) or G4 curvature continuity (tangent angle and curvature plus first and second derivatives of curvature are continuous) is input. The desired curve is expressed as the solution to the chosen set of constraints.

Claims

What is claimed is: 1. A computer-implemented method, comprising: receiving, at a computing device, a first control point and a second control point; determining, at the computing device, a first set of constraints for curvature continuity at the first control point, the first set of constraints including tangent angle and curvature at the first control point; determining, at the computing device, a second set of constraints for curvature continuity at the second control point, the second set of constraints including tangent angle, curvature, first derivative of curvature and second derivative of curvature at the second control point; computing, at the computing device, a curve having a curvature and an arclength at each point along the curve, the curvature at each specific point of the curve being a polynomial function of the arclength corresponding to that specific point, wherein the curve satisfies the first set of constraints at the first control point and wherein the curve satisfies the second set of constraints at the second control point; and outputting, to a display device or printer, the curve. 2. The method of claim 1 , wherein: determining the first set of constraints for curvature continuity at the first control point comprises receiving the first set of constraints; and determining the second set of constraints for curvature continuity at the second control point comprises receiving the second set of constraints. 3. The method of claim 2 , wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point, and wherein the curve satisfies the second set of constraints at the second control point when the tangent angle, curvature, first derivative of curvature and second derivative of curvature are equal across the second control point. 4. The method of claim 2 , wherein the curve satisfies the second set of constraints at the second control point when the first derivative of curvature and second derivative of curvature are zero on one side of the second control point and are non-zero on another side of the second control point. 5. The method of claim 1 , wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point, and wherein the curve satisfies the second set of constraints at the second control point when the tangent angle, curvature, first derivative of curvature and second derivative of curvature are equal across the second control point. 6. The method of claim 1 , wherein the curve satisfies the second set of constraints at the second control point when the first derivative of curvature and second derivative of curvature are zero on one side of the second control point and are non-zero on another side of the second control point. 7. The method of claim 1 , wherein the first control point comprises a transition between a straight section and a curved section of the curve, and the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point. 8. A computing device for creating a character of a font, comprising: a font designing means for receiving a first control point and a second control point, and for determining a first set of constraints for curvature continuity at the first control point and a second set of constraints for curvature continuity at the second control point, the first set of constraints including tangent angle and curvature at the first control point and the second set of constraints including tangent angle, curvature, first derivative of curvature and second derivative of curvature at the second control point; a font generation means for computing a curve having a curvature and an arclength at each point along the curve, the curvature at each specific point of the curve being a polynomial function of the arclength corresponding to that specific point, wherein the curve satisfies the first set of constraints at the first control point and wherein the curve satisfies the second set of constraints at the second control point; and a memory that stores the curve and the first and second control points corresponding to the character. 9. The computing device of claim 8 , wherein: determining the first set of constraints for curvature continuity at the first control point comprises receiving the first set of constraints; and determining the second set of constraints for curvature continuity at the second control point comprises receiving the second set of constraints. 10. The computing device of claim 9 , wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point, and wherein the curve satisfies the second set of constraints at the second control point when the tangent angle, curvature, first derivative of curvature and second derivative of curvature are equal across the second control point. 11. The computing device of claim 9 , wherein the curve satisfies the second set of constraints at the second control point when the first derivative of curvature and second derivative of curvature are zero on one side of the second control point and are non-zero on another side of the second control point. 12. The computing device of claim 8 , wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point, and wherein the curve satisfies the second set of constraints at the second control point when the tangent angle, curvature, first derivative of curvature and second derivative of curvature are equal across the second control point. 13. The computing device of claim 8 , wherein the curve satisfies the second set of constraints at the second control point when the first derivative of curvature and second derivative of curvature are zero on one side of the second control point and are non-zero on another side of the second control point. 14. The computing device of claim 8 , wherein the first control point comprises a transition between a straight section and a curved section of the curve, and wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point. 15. A computer-implemented method, comprising: receiving, at a computing device, a first control point and a second control point; determining, at the computing device, a first set of constraints for curvature continuity at the first control point, the first set of constraints including tangent angle and curvature at the first control point; determining, at the computing device, a second set of constraints for curvature continuity at the second control point, the second set of constraints including tangent angle and curvature at the second control point; and computing, at the computing device, a curve having a curvature and an arclength at each point along the curve, the curvature at each specific point of the curve being a polynomial function of the arclength corresponding to that specific point, wherein the curve satisfies the first set of constraints at the first control point and wherein the curve satisfies the second set of constraints at the second control point. 16. The method of claim 15 , wherein: determining the first set of constraints for curvature continuity at the first control point comprises receiving the first set of constraints; and determining the second set of constraints for curvature continuity at the second control point comprises receiving the second set of constraints. 17. The method of claim 16 , wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point, and wherein the curve satisfies the second set of constraints at the second control point when the tangent angle and curvature are equal across the second control point. 18. The method of claim 16 , wherein the curve satisfies the second set of constraints at the second control point when a first derivative of curvature and second derivative of curvature are zero on one side of the second control point and are non-zero on another side of the second control point. 19. The method of claim 15 , wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point, and wherein the curve satisfies the second set of constraints at the second control point when the tangent angle and curvature are equal across the second control point. 20. The method of claim 15 , wherein the first control point comprises a transition between a straight section and a curved section of the curve, and wherein the curve satisfies the first set of constraints at the first control point when the tangent angle and the curvature are equal across the first control point.
CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Application No. 60/747,931, which was filed on May 22, 2006, and U.S. Provisional Application No. 60/915,919, which was filed on May 3, 2007. The disclosure of each of the above applications are incorporated herein by reference in their entirety. FIELD OF THE INVENTION The present invention relates to interactive curve design. BACKGROUND OF THE INVENTION The literature of curve interpolation techniques, otherwise known as “splines” is vast. A system for interactive design of curves should have the following properties: (a) The system should be capable of accurately representing a rich variety of curves with a minimum of input (measured as the number of control points); (b) The resulting curves should be smooth; (c) When moving control points, the effect on the curve should be direct and simple to understand; and (d) The computational resources should be modest enough to allow for update of the display at interactive rates. Much curve design, including most font design, is done using cubic Bezier curves. FIG. 3 shows a typical cubic Bezier segment with knots and additional control points. In a Bezier curve, some of the control points (“knots or knot points”) lie on the curve ( 30 , 36 ), but others lie off the curve ( 32 , 34 ), and the position of the off-curve control points controls the shape of the curve. As the designer moves control points 32 and 34 , the shape of the curve between control points 30 and 36 changes. This technique is simple to compute and very fast, and is good at representing a rich diversity of curves, but is not very simple to control. It is especially difficult to achieve smooth curves through the on-curve control points, and it takes considerable time and effort for designers to learn to do so. A particularly difficult aspect of curve design for font generation is making smooth transitions between straight and curved sections. Smoothness is a subjective quality, and there is no consensus in the literature on the best way to measure it, but one very commonly related objective measure closely connected to smoothness is the degree of continuity across the knot points. This continuity measure, in turn, is defined in terms of “curvature”. Curvature is defined as the inverse of instantaneous radius, i.e., the inverse of the radius of the osculating circle that “kisses” the curve at every point. If the curvature exists at all points (i.e. there are no sharp corners), then the curve has G1 continuity. If the curvature is also continuous through the control point, then it has G2 continuity. Higher degrees of continuity are defined in terms of derivatives of the curvature with respect to arclength. If the first derivative of curvature is continuous, then the curve has G3 continuity, and if the second derivative of curvature is also continuous, then the curve has G4 continuity. In general, cubic Bezier curves only guarantee G1 continuity. Another class of curve generation techniques involves “interpolating splines,” or curves that go through (“interpolate”) all the control points. One example used in practice for font design is “Smooth, Easy to Compute Interpolating Splines,” by John Hobby, Stanford University, 1985 as implemented in the METAFONT system. In the prior art METAFONT spline, the range of choices for constraints is limited (the constraint is that of continuity of “mock curvature” across control points). In curve design, a primitive is defined as the basic element upon which more complex curves are built. The underlying primitive curve in the METAFONT system is a reduced family of cubic polynomials (with a two-dimensional parameter space). Another prior art system known as Ikarus, by Peter Karow, is similar to METAFONT. Yet another class of interpolating spline techniques is the “variational approach,” in which a curve is chosen to minimize some energy functional. Such techniques emulate mechanical splines based on thin strips of elastic material such as metal or wood. Examples include Gumhold (Designing Optimal Curves in 2D, Proc. CEIG, July 2004 pages 61-76), and Moreton (Minimum Curvature Variation Curves, Networks and Surfaces for Fair Free-Form Shape Design, 1992 University of California, Berkeley Calif.). In general, variational techniques provide for a high degree of smoothness, but also require substantial computational resources. Moreton described MEC (Minimum Energy Curve) with G2 continuity and also MVC (Minimum Variation of Curvature) splines with G4 continuity. SUMMARY OF THE INVENTION The invention is a novel technique for interactive curve design, comprising several innovative features. Although the present invention is illustrated for use in conjunction with font design, the technique may be applied railroad track design, roadway design, and general CAD applications involving smooth curves. In accordance with a first feature, the present invention uses as its primitive, a curve whose curvature is a polynomial function of arclength, i.e. whose intrinsic equation is a polynomial. In accordance with a second feature, for each control point, the present invention allows for a choice of G2 or G4 curvature continuity. In accordance with a third feature, the present invention provides for transitions between straight and curved sections, in which G2 continuity is preserved. In the present system, the desired curve is expressed as the solution to a set of constraints. The set of constraints is chosen from the tangent angle, curvature, first derivative of curvature, and second derivative of curvature. In summary, the input to the system is a sequence of control points and constraint codes for each control point, and the output is a curve in which each of the constraints is satisfied. DESCRIPTION OF THE DRAWING FIG. 1 illustrates the use of the present invention for font generation. FIG. 2 illustrates the use of the present invention for interactive font design. FIG. 3 illustrates a method of curve generation in accordance with the prior art. FIG. 4 is an example of an Euler spiral for use in conjunction with the present invention. FIG. 5 shows a unit-length normalized polynomial spiral segment for use in conjunction with the present invention. FIG. 6A shows a typical font outline generated using the present invention. FIG. 6B shows a typical font outline generated using the present invention, with an Euler spiral overlaying one such segment. FIG. 7 shows a family of font outlines of varying weight in accordance with the present invention, produced by linearly interpolating the control points. FIG. 8 shows the family of font outlines shown in FIG. 7 , superimposed to show more clearly the relations between the curves of the different outlines. FIG. 9A shows a font outline for a “U” letter in accordance with the present invention. FIG. 9B is a plot showing curvature as a function of arclength for the font outline of the “U” letter in FIG. 9A in accordance with the present invention. DETAILED DESCRIPTION OF THE INVENTION A system using the present invention for document printing is illustrated in FIG. 1 . A document 100 includes text 104 and a font identifier (font ID) 102 . The font ID 102 contains the control points (data points or knots) for each character in the chosen font. Prior to printing or displaying the document 100 , a font generator 106 renders each character, which is then stored in a bitmap memory 108 . The present invention for curve generation finds its utility in the font generator 106 where the curves between control points are computed. In order to print (or display) the document 100 , individual characters of the text 104 contained in the document 100 are sent to the print/display engine 110 , which uses the stored bitmap of each character in memory 108 to render the document 100 on a display 112 or generate a printer output 114 . An interactive system using the present invention for font design is illustrated in FIG. 2 . The font designer provides input 200 in the form of selecting control points 201 that correspond to the shape of the desired font characters. The control points are sent to the font generator 206 . In accordance with the present invention, the font generator 206 computes the curves between the selected control points 201 . The resulting bitmap for such character is stored in a bitmap memory 208 . A print/display engine 210 generates a display 212 and/or a printed output 214 for evaluation by the font designer. The rapid computation of curves by the font generator 206 on the display 212 permits effective interactive font design. When the font designer is satisfied with the shape of the displayed character, the control points are saved 204 in a font ID memory 202 . The interactive qualities achieved by the use of the present invention are illustrated in FIG. 7 and FIG. 8 . FIG. 7 shows the result using the present invention when the control points are linearly interpolated between two versions of a font outline. In particular, FIG. 7 shows a thin version 7 A of a font linearly interpolated up to a thick version 7 K of the same font. All such instances of the font 7 A, 7 B . . . , 7 J, 7 K, have G2 continuity everywhere (other than corners). By contrast, linearly interpolating Bezier control points, as in the Adobe Multiple Master technology, does not even guarantee G1 continuity in the general case. FIG. 8 illustrates the same linear interpolation between a thin version 8 A of a font linearly interpolated up to a thick version 8 K of the same font. In particular, FIG. 8 shows a thin version 8 A of a font linearly interpolated up to a thick version 8 K of the same font superimposed upon each other. All such instances of the font 8 A, 8 B . . . , 8 J, 8 K, have G2 continuity everywhere (other than corners). The Primitive Curve The preferred primitive curve of the present invention is a third-order polynomial spiral. A polynomial spiral is a curve whose curvature expressed in terms of arclength (its “intrinsic equation”) is given by a polynomial. Each such curve is characterized by four scalar parameters, which may be represented as the curvature and its first three derivatives at the midpoint of the curve. A few special cases are worth noting. If all four parameters are zero, then the curve is a straight line. If the curvature is nonzero but its derivatives are zero, then the curve is a circular arc of constant curvature. If the first two parameters are nonzero and the second and third derivatives of curvature are zero, then the curve is a clothoid, or a segment of the spiral of Cornu. Polynomial spirals are also known as “generalized clothoids”. The Constraints Since each segment is a determined by four parameters, in a closed curve each point must have exactly four scalar constraints, to make sure that the resulting curve is fully determined. A G4 constraint specifies that the tangent angle, curvature, first derivative of curvature, and second derivative of curvature must be equal across the control point. A G2 constraint specifies that the tangent angle and curvature must be equal across the control point. To satisfy the requirement that each control point has a total of four scalar constraints, we add two additional constraints: the second derivative of curvature must be zero on both sides of the control point. A one-way constraint with the straight side on the left and the curved side on the right is as follows. As in a G2 constraint, tangent angle and curvature must be equal across the control point. On the left (straight side), both first and second derivative of curvature must be zero. The palette of constraints contains eight possible choices, namely the tangent angle, curvature, first derivative of curvature, and second derivative of curvature on one side of the control point, and the tangent angle, curvature, first derivative of curvature, and second derivative of curvature on the other side of the control point. In general, constraints at each control point can be written in terms of a linear combination of these eight possible parameters set equal to a constant (typically zero). Any set of four of these constraints is a valid set of constraints. In an open curve, there is one more point than the number of segments, so the endpoints are treated specially. Each endpoint is assigned two constraints, specifically that the first and second derivatives of curvature are zero. In this way, the total number of constraints is always four times the number of segments, as required. These constraints can be mixed freely and at will. The resulting curve, after solution of the spline, has a number of notable properties. First, any segment bounded by a standard G2 constraint on both ends is a segment of an Euler spiral, also known as clothoid. A spline consisting entirely of G2 constraints is known as an Euler spiral spline. Euler's spiral, plotted in FIG. 4 over the range [−5, 5], is characterized by its intrinsic definition: curvature is linear with arclength, namely, k(s)=s/2. Second, if all points have G4 control points, the spline described by Ohlin in “Splines for Engineers.” spline is generated. Third, if points are perturbed on the curved side of a one-way constraint, the curve on the straight side is not affected. In addition to providing for smooth straight-to-curved transitions, this property allows the construction of curves more robust with respect to perturbation, as may be useful when the control points themselves are parameterized or interpolated. Implementation of the Constraint Solver The core of the preferred embodiment of the present invention is a constraint solver specialized to the particular systems of constraints that will be solved: a sequence of segments each of which is a polynomial spiral controlled by four parameters, and an equal number of constraints written in terms of the tangent angles, curvatures, and derivatives of curvature at the endpoints of each of the segments. In broad outline, the solver works as follows: 1. Determine initial values for all the parameters. One reasonable choice is all zeros, so that initially the segments are straight lines. The result of this step can be represented as a vector. 2. Evaluate the tangent angles, curvatures, and derivatives of curvature at the endpoints of each of the segments. 3. For each constraint, evaluate a quantity that represents the deviation from the constraint. For constraints of the form “quantity equals zero”, this deviation is the quantity itself. For constraints of the form “first quantity=second quantity”, this deviation is the difference between the two quantities, i.e. “first quantity minus second quantity”. The result of this step is a vector of deviation values. 4. Determine whether the deviation vector is within an acceptable tolerance of zero. If it is exactly zero, then the constraints are exactly satisfied. If it is close to zero, then the constraints are approximately satisfied. 5. If the deviation vector is within tolerance, then the solver is complete, and the result is the vector of parameters. 6. For each combination of parameter value and deviation value, compute an approximate value of how an infinitesimal change to the parameter will affect the deviation. The result of this step is a matrix. Further, this matrix is sparse and can be represented in band-diagonal form because changes to parameters only affect the deviations at the left and right endpoints. 7. Invert the matrix generated in step 6, above. While inversion of arbitrary matrices is fairly expensive computationally, efficient algorithms exist for inverting matrices in band-diagonal form. See Numerical Recipes in C, 2 nd ed. by William H. Press, Brian P. Flannery, Saul A. Teukolsky, William T. Vetterling. Cambridge University Press, 1992. 8. Compute the dot product of the inverted matrix and the deviation vector, resulting in an update vector. 9. Subtract the update vector from the parameter vector, resulting in a refined parameter vector. 10. Go back to step 2 and repeat the process. Evaluation of Tangent Angles, Curvatures, and Derivatives of Curvature A central sub-problem of the constraint solver (step 2 of the above section) is to evaluate the tangent angles, curvatures, and derivatives at each of the two endpoints (control points) of each polynomial spiral segment, given the four parameters of the polynomial spiral. The standard representation of a third-order polynomial spiral segment is as a vector of four quantities measured at the arclength midpoint of the segment normalized to unit length: curvature, and the first three derivatives of curvature. By convention, the midpoint has Cartesian coordinates at the origin (0,0), and has a tangent angle of zero, i.e. parallel to the horizontal axis. FIG. 5 shows an example of such a segment 50 between a first and second endpoints 52 , 54 . Using a Taylor's series expansion, the curvature of segment 50 measured at arclength s from this central point, (i.e. s ranges from −0.5 to 0.5), is: k ( s )= k 0 +k 1 s+ ½ k 2 s^ 2+⅙ k 3 s^ 3 The tangent angle at each point along the curve 50 , measured in terms of the arclength parameter s, can be determined by simple integration of this curvature value: theta( s )= k 0 s+ ½ k 1 s^ 2+⅙ k 2 s^ 3+ 1/24 k 3 s^ 4 The Cartesian coordinates of each point along the curve 50 , measured in terms of the arclength parameter s, can be determined by numerical integration of the sine and cosine of the tangent angle. x ( s )=integral (from 0)(to s )cos theta( t ) dt y ( s )=integral (from 0)(to s )sin theta( t ) dt The chordlength of this segment, which is equal to the ratio of chordlength divided by arclength because the arclength is unity, is simply the Pythagorean distance between the two endpoints, represented in Cartesian coordinates as (x(−0.5), y(−0.5)) and (x(0.5), y(0.5)), respectively. Thus, this value is sqrt((y(0.5)−y(−0.5))^2+((x(0.5)−x(−0.5))^2)). The unit-length normalized segment is then rotated and scaled so that its endpoints match the control points given as input to the constraint solving problem. The scale factor is equal to the chordlength between the control points divided by the chordlength of the unit-length normalized segment. Similarly, the rotation angle is equal to the angle of the vector from the left control point to the right control point, minus the angle of the vector from the left endpoint to the right endpoint of the unit-length normalized segment (the latter angle is marked by the Greek letter psi in FIG. 5 ). With these values (the scale factor and rotation angle), the tangent angles, curvatures, and derivatives can be readily calculated for all points along the curve, including the two endpoints. tangent angle( s )=(rotation angle)+theta( s )=(rotation angle)+ k 0 s+ ½ k 1 s^ 2+⅙ k 2 s^ 3+ 1/24 k 3 s^ 4 curvature( s )= k ( s )/(scale factor)^2=( k 0 +k 1 s+ ½ k 2 s^ 2+⅙ k 3 s^ 3)/(scale factor) curvature′( s )= k′ ( s )/(scale factor)^2=( k 1 +k 2 s+ ½ k 3 s^ 2)/(scale factor)^2 curvature″( s )= k′ ( s )/(scale factor)^3=( k 2 +k 3 s )/(scale factor)^3 These four formulae are then evaluated at s=−0.5 and s=0.5, resulting in the values for the left and right endpoint, respectively. FIG. 6A illustrates an “a” letter 64 generated for a typical font outline. Although the curve generation in accordance with the present invention determines only the outline of the “a” letter 64 , it will be understood that when printing and displaying the letter outline 64 , the interior of the curve will be a solid color (typically black). The control points such as 60 , 61 , 62 and 63 in FIG. 6A identify the font outline. In accordance with the present invention, a portion of a Euler spiral 640 is fitted between control points 60 and 63 in FIG. 6B . The power of the present invented algorithm for fitting curves between control points is illustrated by the fact that a single Euler spiral 640 passes through all four control points 60 , 61 , 62 and 63 . FIG. 9A illustrates a font outline for an asymmetrical “U” letter 90 . The plot in FIG. 9B shows curvature as a function of arclength for the font outline of the “U” letter 90 in FIG. 9A . The peak curvature points 92 , 94 , 96 , 98 correspond to the corresponding curve segments 92 , 94 , 96 , 98 in FIG. 9A . In between, the short straight sections 93 and 97 in FIG. 9A correspond to the zero curvature segments 93 and 97 in FIG. 9B . Similarly, the large curved section 91 and the small curved section 95 in FIG. 9A correspond to the positive curvature 91 and negative curvature 95 in FIG. 9B . Note the smooth transitions between straight sections (such as segment 99 ) and curve segments (such as segment 92 ).

Description

Topics

Download Full PDF Version (Non-Commercial Use)

Patent Citations (21)

    Publication numberPublication dateAssigneeTitle
    US-2003120427-A1June 26, 2003Chapman Clinton D., Jie ZhangGraphical method for designing the trajectory of a well bore
    US-2005071021-A1March 31, 2005Rockwell Automation Technologies, Inc.Method and system for generating multi-dimensional motion profiles
    US-2005106998-A1May 19, 2005Wen-Jong Lin, Ng Tsong J., Chen Xiao Q., Zhiming GongMethod of determining shape data
    US-2005175238-A1August 11, 2005Sebastien RoselProcess for drafting a curve in a computer-aided design system
    US-2005190187-A1September 01, 2005Microsoft CorporationMethods and systems for hinting fonts
    US-2005237342-A1October 27, 2005Beat Stamm, Hitchcock Gregory C, Claude Betrisey, Matt ConwayAutomatic optimization of the posistion of stems of text characters
    US-2007116339-A1May 24, 2007Siemens Corporate Research IncSystem and Method For Myocardium Segmentation In Realtime Cardiac MR Data
    US-5081594-AJanuary 14, 1992Kroy Inc.Real-time rasterization system for a column-oriented printing apparatus or the like
    US-5235551-AAugust 10, 1993Pacific Data Products, Inc.Memory addressing scheme
    US-5295238-AMarch 15, 1994Highlights For Children, Inc.System, method, and font for printing cursive character strings
    US-5471574-ANovember 28, 1995Xerox CorporationMethod for displaying a computer generated graphic on a raster output scanner
    US-5636338-AJune 03, 1997Silicon Graphics, Inc.Method for designing curved shapes for use by a computer
    US-5731820-AMarch 24, 1998Novell, Inc.Curve-length approximation apparatus and method
    US-5781714-AJuly 14, 1998Bitstream Inc.Apparatus and methods for creating and using portable fonts
    US-5949435-ASeptember 07, 1999Adobe Systems IncorporatedDigital type font providing typographic feature transformation capability
    US-6014148-AJanuary 11, 2000Laser Products, Inc.Method for generating two dimensional and three dimensional smooth curves and for driving curve forming devices
    US-6441823-B1August 27, 2002Brigit AnanyaComputer curve construction system
    US-6459439-B1October 01, 2002Macromedia, Inc.Reshaping of paths without respect to control points
    US-6636217-B1October 21, 2003Autodesk, Inc.Regularized tangents in computer graphics
    US-6639592-B1October 28, 2003Silicon Graphics, Inc.Curve network modeling
    US-7057616-B2June 06, 2006Microsoft CorporationUsing constrained optimization in curve editing

NO-Patent Citations (5)

    Title
    Gumhold, "Designing Optimal Curves in 2D", Jul. 2004, Spain.
    Hobby, "Smooth, Easy to Compute Interpolating Splintes", 1985, Stanford University, Palo Alto, CA.
    Knuth, "The Metafont Book", 1986, pp. 122-139, Addison-Wesley, Boston, MA.
    Moreton, "Minimum Curvature Variation Curves, Networks, and Surfaces for Fair Free-Form Shape Design", 1992, University of California, Berkley, Berkley, CA.
    Ohlin, "Splines for Engineers", Eurographics, 1987, pp. 555-565, Elservier Science Publishers, Holland.

Cited By (0)

    Publication numberPublication dateAssigneeTitle