As I am currently learning Python, I thought that this would be a good programming exercise. I Googled for a while and decided on using a rose curve.
After playing around a while with the output in Photoshop, I settled on what's currently in the background.
If you're curious, the Python code is below. To generate it image, I took the rose curve from the code below, layered it on itself many times, adjusted the color, applied some Gaussian blur, and then resized.
import numpy as np
###############################################################
def rose1(angle,a,n,d):
r=1+(a*np.cos((n/d)*angle))
return(r)
#############################################################
print("\nEquation is of the form:")
print("For phi in (0,2*D*pi): r=1+(A*(cos((N/D)phi))")
print("N and D should be relatively prime whole numbers.")
A=float(raw_input("\nWhat do you want A to be? "))
N=float(raw_input("What do you want N to be? "))
D=float(raw_input("What do you want D to be? "))
theta=np.arange(0,2*D*3.14159265359,0.005)
r=rose1(theta,A,N,D)
x=r*np.cos(theta)
y=r*np.sin(theta)
fig1=plt.figure(figsize=(15,15))
sub1=fig1.add_subplot(111,frameon=False,aspect='equal')
sub1.plot(x,y)
sub1.tick_params(axis='both',which='both',length=0, width=0, labelbottom=False,labeltop=False,labelleft=False,labelright=False)
plt.show()
This code creates a rose curve without any axes or grid marks, making it much easier to copy into Photoshop (sample output below):
All of this can probably be coded much more efficiently, but this version worked for my purposes.