minkowski

Modules

Module	Description
modeling/minkowski	Minkowski sum operations for 3D geometries.

Functions

Name	Description
minkowskiSumNonConvexConvex()	Compute Minkowski sum of non-convex A with convex B.
decomposeIntoTetrahedra()	Decompose a geom3 into tetrahedra using face-local apex points. Each resulting tetrahedron is guaranteed to be convex.
createTetrahedronPolygons()	Create the 4 triangular faces of a tetrahedron.
minkowskiSumConvex()	Compute Minkowski sum of two convex polyhedra.
extractUniqueVertices()	Extract unique vertices from a geom3. Uses a Set with string keys for deduplication.

modeling/minkowski

Minkowski sum operations for 3D geometries. 两个形状 A 与 B 的闵可夫斯基和，是所有由 A 中一点与 B 中一点相加得到的点的集合。该运算常用于：

• 形状的偏移 / 膨胀（使用球体可生成圆角边）
• 碰撞检测（当且仅当两个形状的闵可夫斯基差包含原点时，它们发生碰撞）
• 运动规划与扫掠体积计算

Example

const { minkowskiSum } = require('@jscad/modeling').minkowski
const rounded = minkowskiSum(cube, sphere)

modeling/minkowski.minkowskiSum(...geometries)

计算两个三维几何体的闵可夫斯基和。

闵可夫斯基和 A⊕B 是所有满足 a+b 的点的集合，其中 a∈A，b∈B。 从几何上看，这相当于用几何体 B 的形状 “膨胀” 几何体 A。

Common use cases:

• Offset a solid by a sphere to round all edges and corners
• Offset a solid by a cube to create chamfered edges
• Collision detection (if Minkowski sum contains origin, shapes overlap)

为获得最佳性能，请使用凸几何体。当第二个操作数是凸几何体时，也支持非凸几何体，但需要进行分解处理，速度会更慢。

Kind: static method of modeling/minkowski
Returns: geom3 - new 3D geometry representing the Minkowski sum

Param	Type	Description
...geometries	Object	two geom3 geometries (second should be convex for non-convex first)

Example

const { primitives, minkowski } = require('@jscad/modeling')
const cube = primitives.cuboid({ size: [10, 10, 10] })
const sphere = primitives.sphere({ radius: 2, segments: 16 })
const rounded = minkowski.minkowskiSum(cube, sphere)

minkowskiSumNonConvexConvex()

Compute Minkowski sum of non-convex A with convex B.

Decomposes A into tetrahedra, computes Minkowski sum of each with B, then unions all results.

Kind: global function

decomposeIntoTetrahedra()

Decompose a geom3 into tetrahedra using face-local apex points. Each resulting tetrahedron is guaranteed to be convex.

Unlike centroid-based decomposition, this approach works correctly for shapes where the centroid is outside the geometry (e.g., torus, U-shapes). Each polygon gets its own apex point, offset inward along its normal.

Kind: global function

createTetrahedronPolygons()

Create the 4 triangular faces of a tetrahedron.

Kind: global function

minkowskiSumConvex()

Compute Minkowski sum of two convex polyhedra.

For convex polyhedra, the Minkowski sum equals the convex hull of all pairwise vertex sums. This is O(n*m) for n and m vertices, plus the cost of the convex hull algorithm.

Kind: global function

extractUniqueVertices()

Extract unique vertices from a geom3. Uses a Set with string keys for deduplication.

Kind: global function