OptAtlas
Formal problem

2D Knapsack

Select a subset of rectangles for a single sheet, maximizing value (or area).

Also called: 2D Knapsack · Rectangle knapsack · 직사각형 배낭

Last verified: 2026-05-27

Definition

Choose a subset of candidate rectangles and place them, without overlap, on a single fixed-size sheet, maximizing the total value (often area) of the placed pieces.

Example

Suppose a sheet 4×44\times4 and candidates 3×33\times3 (value 9), 4×14\times1 (value 4), 1×31\times3 (value 3), and 2×22\times2 (value 4). Choosing the first three — 3×33\times3 at the bottom-left, 4×14\times1 along the top row, 1×31\times3 in the right column — tiles the sheet exactly for value 9+4+3=169+4+3=16. Adding the 2×22\times2 would exceed the sheet area, so this trio is optimal.

Family

A core orthogonal cutting & packing problem alongside bin packing and strip packing. Its "what to include" selection structure on a single sheet shares its name and structure with the one-dimensional 0/1 Knapsack Problem.

Related nodes

See the depth-1 graph below.

Claims & evidence

Every relationship is a claim with an equivalence level and an evidence grade. See the evidence policy.

RelationshipClaimEquiv.EvidenceSources
uses methodDynamic Programming2D knapsack variants have been addressed with dynamic programming and exact/approximate methods built on it.B
  • AKnapsack Problems
direct benchmark2DPackLib2DPackLib includes two-dimensional orthogonal knapsack instances.A
  • A2DPackLib: a two-dimensional cutting and packing library
shares method with2D Bin Packing2D knapsack and bin packing share the orthogonal placement core and many exact/heuristic methods.E2B
  • AAn improved typology of cutting and packing problems
generalizes0/1 Knapsack Problem2D knapsack is a two-dimensional generalization of the 1D 0/1 knapsack problem, extending item selection to geometric placement.E2B
  • AKnapsack Problems
uses methodBranch and BoundExact branch-and-bound algorithms have been proposed for the geometric 2D knapsack, using a one-dimensional knapsack relaxation (rectangle areas as weights) as a bound (Caprara & Monaci 2004).A
  • AOn the two-dimensional Knapsack Problem
uses methodGenetic Algorithm2D knapsack variants have also been addressed with metaheuristics such as genetic algorithms.B
  • AAn improved typology of cutting and packing problems

Neighborhood

Direct graph neighbors. Toggle depth to expand.

Click a node to open it · click an edge for its claim

See also

Not directly linked, but conceptually close — by the connections and descriptions they share.