Java implementation of Perlin Noise algorithm as described in this resource.
This project is an implementation of the popular Perlin noise algorithm described by Ken Perlin.
This projects also features generators to populate grids of noise in 1 and 2 dimensions that can be updated over time. These generators offer the option of spatial circularity over the grid's borders. For example, connecting both ends of a line generated this way will preserve noise continuity over the new-made junction. For a 2D grid, tiling the grid with itself would also preserve noise continuity over the tile's junction.
The perlin noise core algorithm is accessible through the PerlinNoise class. After creating an instance of the class with a random seed, it can be used as follows:
PerlinNoise perlinNoise = new PerlinNoise(seed);
double noiseValue = perlinNoise.getFor(coordinates);
where coordinates represents a single dimension array of doubles, of length up to 5.
The noise generators aim to automatize the generation of noise following a single or double dimensional array. The generators are accessible through their respective builders like such:
LayeredSliceGeneratorBuilder builder = new LayeredSliceGeneratorBuilder(width, height);
LayeredSliceGenerator generator = builder.build();
double[][] noiseValues = generator.getNext();
The number of layers is set using the builder:
builder.withNumberOfLayers(n);
Step sizes in spatial and time dimensions are set with the builder's methods:
builder.setTimeStepSizes(timeStepSizes);
builder.setWidthStepSizes(widthStepSizes);
builder.setHeightStepSizes(heightStepSizes);
where stepSizes is an Iterable<Double> having at least n (number of layers) values.
The amplitude of each layer is set with the builder's method:
builder.withAmplitudes(amplitudes);
which is also an Iterable<Double>, following the same principle as the setXStepSizes() methods.
Spatial circularity option is set with the builder:
builder.withCircularBounds(true);
Parallelization of the noise generation is possible using two paradigms:
- the ForkJoinPool Java framework, to split line or slice generation into smaller fragments;
- the ExecutorService framework, to launch generation of each layer simultaneously.
These optimizations are accessible using the builder's methods:
builder.withForkJoinPool(pool);
builder.withLayerExecutorService(executorService);