The Algorithms logo
The Algorithms
关于捐赠

Perlin Noise

d
package com.thealgorithms.others;

import java.util.Random;
import java.util.Scanner;

/**
 * For detailed info and implementation see: <a
 * href="http://devmag.org.za/2009/04/25/perlin-noise/">Perlin-Noise</a>
 */
public final class PerlinNoise {
    private PerlinNoise() {
    }

    /**
     * @param width width of noise array
     * @param height height of noise array
     * @param octaveCount numbers of layers used for blending noise
     * @param persistence value of impact each layer get while blending
     * @param seed used for randomizer
     * @return float array containing calculated "Perlin-Noise" values
     */
    static float[][] generatePerlinNoise(int width, int height, int octaveCount, float persistence, long seed) {
        final float[][] base = new float[width][height];
        final float[][] perlinNoise = new float[width][height];
        final float[][][] noiseLayers = new float[octaveCount][][];

        Random random = new Random(seed);
        // fill base array with random values as base for noise
        for (int x = 0; x < width; x++) {
            for (int y = 0; y < height; y++) {
                base[x][y] = random.nextFloat();
            }
        }

        // calculate octaves with different roughness
        for (int octave = 0; octave < octaveCount; octave++) {
            noiseLayers[octave] = generatePerlinNoiseLayer(base, width, height, octave);
        }

        float amplitude = 1f;
        float totalAmplitude = 0f;

        // calculate perlin noise by blending each layer together with specific persistence
        for (int octave = octaveCount - 1; octave >= 0; octave--) {
            amplitude *= persistence;
            totalAmplitude += amplitude;

            for (int x = 0; x < width; x++) {
                for (int y = 0; y < height; y++) {
                    // adding each value of the noise layer to the noise
                    // by increasing amplitude the rougher noises will have more impact
                    perlinNoise[x][y] += noiseLayers[octave][x][y] * amplitude;
                }
            }
        }

        // normalize values so that they stay between 0..1
        for (int x = 0; x < width; x++) {
            for (int y = 0; y < height; y++) {
                perlinNoise[x][y] /= totalAmplitude;
            }
        }

        return perlinNoise;
    }

    /**
     * @param base base random float array
     * @param width width of noise array
     * @param height height of noise array
     * @param octave current layer
     * @return float array containing calculated "Perlin-Noise-Layer" values
     */
    static float[][] generatePerlinNoiseLayer(float[][] base, int width, int height, int octave) {
        float[][] perlinNoiseLayer = new float[width][height];

        // calculate period (wavelength) for different shapes
        int period = 1 << octave; // 2^k
        float frequency = 1f / period; // 1/2^k

        for (int x = 0; x < width; x++) {
            // calculates the horizontal sampling indices
            int x0 = (x / period) * period;
            int x1 = (x0 + period) % width;
            float horizintalBlend = (x - x0) * frequency;

            for (int y = 0; y < height; y++) {
                // calculates the vertical sampling indices
                int y0 = (y / period) * period;
                int y1 = (y0 + period) % height;
                float verticalBlend = (y - y0) * frequency;

                // blend top corners
                float top = interpolate(base[x0][y0], base[x1][y0], horizintalBlend);

                // blend bottom corners
                float bottom = interpolate(base[x0][y1], base[x1][y1], horizintalBlend);

                // blend top and bottom interpolation to get the final blend value for this cell
                perlinNoiseLayer[x][y] = interpolate(top, bottom, verticalBlend);
            }
        }

        return perlinNoiseLayer;
    }

    /**
     * @param a value of point a
     * @param b value of point b
     * @param alpha determine which value has more impact (closer to 0 -> a,
     * closer to 1 -> b)
     * @return interpolated value
     */
    static float interpolate(float a, float b, float alpha) {
        return a * (1 - alpha) + alpha * b;
    }

    public static void main(String[] args) {
        Scanner in = new Scanner(System.in);

        final int width;
        final int height;
        final int octaveCount;
        final float persistence;
        final long seed;
        final String charset;
        final float[][] perlinNoise;

        System.out.println("Width (int): ");
        width = in.nextInt();

        System.out.println("Height (int): ");
        height = in.nextInt();

        System.out.println("Octave count (int): ");
        octaveCount = in.nextInt();

        System.out.println("Persistence (float): ");
        persistence = in.nextFloat();

        System.out.println("Seed (long): ");
        seed = in.nextLong();

        System.out.println("Charset (String): ");
        charset = in.next();

        perlinNoise = generatePerlinNoise(width, height, octaveCount, persistence, seed);
        final char[] chars = charset.toCharArray();
        final int length = chars.length;
        final float step = 1f / length;
        // output based on charset
        for (int x = 0; x < width; x++) {
            for (int y = 0; y < height; y++) {
                float value = step;
                float noiseValue = perlinNoise[x][y];

                for (char c : chars) {
                    if (noiseValue <= value) {
                        System.out.print(c);
                        break;
                    }

                    value += step;
                }
            }

            System.out.println();
        }
        in.close();
    }
}