Perlin 雜訊與傅立葉轉換

perlin_fft.py

import numpy as np
from math import floor
from random import randint
import matplotlib.pyplot as plt

def blending(t):
    return 6 * (t ** 5) - 15 * (t ** 4) + 10 * (t ** 3)
blending = np.frompyfunc(blending, 1, 1)

def lerp(s1, s2, t):
    return s1 + t * (s2 - s1)
lerp = np.frompyfunc(lerp, 3, 1)

def grad1(n, s):
    return s if n % 2 == 0 else -s
grad1 = np.frompyfunc(grad1, 2, 1)

np.random.seed(10) # 為了每次取得相同雜訊，這邊設定了亂數種子
rand_table = np.random.randint(255, size = 256)
def perlin1(x):
    xi = np.floor(x)
    s = x - xi
    a = rand_table[(xi % 256).astype(np.int)]     
    b = rand_table[((xi + 1) % 256).astype(np.int)] 
    return lerp(grad1(a, s), grad1(b, s - 1), blending(s)) # 全部的雜訊

def singal(t, sample_rate):
    noise_level = 10  # 雜訊的變化程度
    x = np.linspace(0, noise_level * t, int(t * sample_rate), endpoint = False)
    return perlin1(x)

t = 2
sample_rate = 800

samples = singal(t, sample_rate)
x = np.linspace(0, t, int(t * sample_rate), endpoint = False)

plt.title('Perlin noise') 
plt.xlabel('x') 
plt.ylabel('y')
plt.gca().set_aspect(1)
plt.plot(x, samples)  
plt.show()