https://github.com/whitegreen/dalsoo-bin-packing
Bin packing places a given set of polygons in standard single/multiple rectangular sheet(s), to minimize the use of the sheet(s).
https://github.com/whitegreen/dalsoo-bin-packing
bin-packing cutting-stock packing-algorithm polygon
Last synced: 8 months ago
JSON representation
Bin packing places a given set of polygons in standard single/multiple rectangular sheet(s), to minimize the use of the sheet(s).
- Host: GitHub
- URL: https://github.com/whitegreen/dalsoo-bin-packing
- Owner: whitegreen
- License: mit
- Created: 2018-10-31T06:37:47.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2023-02-08T02:50:56.000Z (over 3 years ago)
- Last Synced: 2025-04-06T19:40:24.915Z (about 1 year ago)
- Topics: bin-packing, cutting-stock, packing-algorithm, polygon
- Language: Java
- Size: 1.19 MB
- Stars: 58
- Watchers: 4
- Forks: 11
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Dalsoo-Bin-Packing
A java library for 2D irregular shape packing in rectangular sheets. Algorithms are adapted from papers of Abeysooriya 2018 and Dalalah 2014. Bin packing places a given set of polygons in standard single/multiple rectangular sheet(s), to minimize the use of the sheet(s).
This library does not involve other libraries, however, the example uses core.jar (https://processing.org) for a graphical interface. The algorithm is effective when the ratio (number of polygons / number of the types of polygons) is small.
The processing user needs to create “…/libraries/dalsoo/library/dalsoo.jar”, and then write “import pack.*;” to import the library.
## 1. Input
**a**. Only simple polygon: no holes, no self-intersection.
**b**. Data structure: point - double[]; polygon - double[][]; all polygons - double[][][].
One might use other library to convert other formats (e.g., dxf, obj) of polygon to double[][].
**c**. It's better to represent a polygon with a moderate number of points.
Too many points (e.g. a local detail contains dozens of points) slows down the algorithm. Avoid using too many points for a smooth curve.
One might use *segment_max_length* to let the algorithm create more points on a long straight edge of a polygon, if the polygon has very few points or the edge is very long.
## 2. Choose an algorithm
**a**. *useAbey*=true, *Jostle heuristics for the 2D-irregular shapes bin packing problems with free rotation, R. P. Abeysooriya et al. 2018*. However, Jostle method is not used.
*rotSteps*: Each polygon is rotated in the layout. Few steps (say 16) -> run fast & poor result; many steps (say 48) -> run slow & good results
*segment_max_length*: if (an edge of a polygon > segment_max_length), breaks it into smaller segments.
large value -> run fast & poor result; small value -> run slow & good results
*segment_max_length* is related to translation steps.
**b**. *useAbey*=false, *Waste minimization in irregular stock cutting, D. Dalalah et al. 2014*
the rotation/translation steps depend on the polygons.
## 3. Output options
**a**. obtain the transformation of each polygon in the final layout, including:
int[] *result_pack_id*; // result_pack_id[8]=2 means the 9th polygon is on the 3rd sheet.
double[][] *result_cos_sin*; // result_cos_sin[8]={0.5, 0.866} means the 9th polygon is rotated 60 degree around (0,0).
double[][] *result_position*; // result_position[8]={7.4, 9.0} means that the 9th polygon should be translated by (7,4, 9.4) after rotation.
**b**. obtain the geometry of each polygon in the final layout
Pack pack = packs.get(id); //obtain a sheet (pack) by id.
for (Strip strip : pack.fixs) { // obtain a placed strip (polygon) in a sheet.
strip.inps // the polygon's points, one might use other library to convert double[][] to other formats (e.g., dxf: https://github.com/whitegreen/DXF-Laser) of polygon
}
