GMU:Bots 'n' Plots/Apasri Titatarn: Difference between revisions

From Medien Wiki
No edit summary
No edit summary
 
(One intermediate revision by the same user not shown)
Line 29: Line 29:
----Cloud Code----
----Cloud Code----
<pre style="font-size:smaller" >
<pre style="font-size:smaller" >
<source lang="python">
#cloud
#cloud
width = 400
width = 400
Line 83: Line 84:


</pre>
</pre>
</source>




Line 107: Line 109:
[[File:Screen Shot 2558-06-17 at 12.03.18 AM.png]]
[[File:Screen Shot 2558-06-17 at 12.03.18 AM.png]]


As can be seen here,in this step, the final resault take only small part of the source image.
As can be seen here,in this step, the final result take only small part of the source image.


It is quite difficult to trance back what the sorce image look like but it is the preparation for the next step.
It is quite difficult to trace back what the source image look like but it is the preparation for the next step.




code for the image processing
code for the image processing
=============================
=============================
<pre style="font-size:normal" >
 
<source lang="python">


from PIL import Image, ImageDraw,ImageFilter
from PIL import Image, ImageDraw,ImageFilter
Line 220: Line 223:




</pre>
</source>
 




'''FINAL BOT'''


'''STEP 2 : moving Kaleido'''  
'''STEP 2 : moving Kaleido'''  

Latest revision as of 10:03, 16 October 2015

APASRI TITATARN : Bots n Plots class

First touch with Python . . . .gettin to know how to draw basic shapes

Here is my first drawing in class with coding. my sheep robot. . . funny one Sheep.png


My Aim about this class :

" I would like to make imageBot and furthur step would be to make it move as short animated GIF"


Homework 2nd week Assignment to make your own moving robot.

This is my CloudBot. A cloud with falling snowflakes GIF. I coded in processing with python language. Here is the code . . .

Cloud bot s.gif



Cloud Code----

<source lang="python">
#cloud
width = 400
posx = [width/4,width/3,3*width/5,width/2,width/5,2*width/3,3.8*width/5,width/2.3]

def setup():
    size(400,400)
    background(255)
    frameRate(12)
   
    
    
def draw():
    noStroke()
    background(179,226,228)
    change= sin(frameCount*0.1)
    drawcloud(change)
    sx= width/2
    n = frameCount
    textsnow(posx,change)
  
    saveFrame("cloudbot##.jpg")

    

    
def drawcloud(change):
    print change
    fluffy = change*5
    fill(255)
    #cloud 1 S
    ellipse(1.25*width/5,0.9*height/1.8,80+fluffy,80+fluffy)
    #cloud 2 M
    ellipse(2*width/5,height/1.8-height/10,130+2*fluffy,120+1.5*fluffy)
    #cloud 3 L
    ellipse(2.7*width/5,0.95*height/1.8-height/10,120+fluffy,145+fluffy)
    #cloud 4 M
    ellipse(3.5*width/5,height/1.8-height/10,110+2.5*fluffy,110+2*fluffy)
    
    #face
    fill(80)
    ellipse(width/3,height/2,7,7)
    ellipse(2*width/3,height/2,7,7)
    
   
    
def textsnow(posx,change):
    fill(255)
    textSize(32)
    i=0
    while i <7:
        text("*",posx[i]+change*i,height/1.8+frameCount*(i+1)/5)
        i =i+1

</source>



TwitterBot :MID TERM ASSIGNMENT

======Kaleidoscope Bot=======

STEP 1: Still Kaleido

My midtern project is making a image processing bot which make a kaleidoscope style image out of an image. I want the result to look nice and can totally show the kaleidoscope style whatever the source image is.

This is the structure of how I 'kaleidify' the source image.

Kaleidochart1.jpg


RESULT PICTURE

Screen Shot 2558-06-16 at 7.43.39 PM.png Screen Shot 2558-06-17 at 12.03.18 AM.png

As can be seen here,in this step, the final result take only small part of the source image.

It is quite difficult to trace back what the source image look like but it is the preparation for the next step.


code for the image processing

=================
from PIL import Image, ImageDraw,ImageFilter
import PIL.ImageOps as im
import numpy as np

def makekaleido():
#put image path
    imgfile = "img.jpg"

#load Image
    source = Image.open(imgfile)

#adjust the source image a bit for nicer result
    base = source.rotate(31)


#creat mask base the same size with source image 
    mask = Image.new('RGBA', base.size, (255,255,255,0))
    x, y = base.size

# one variable to vary the size of triangle
    var = y/14 

# Define triangle mask position (triangle with 20 degree)
    (originx,originy) = (x/3,int(0.9*y))
    trih = int(12*var)  #fix formular for triangle height
    triw = int(4.2*var) #fix formular for triangle width
    polygonpos = [(originx,originy),
    (originx+triw,originy), 
    (originx+triw/2,originy-trih)]
#print(trih,triw)


# Create mask
    draw = ImageDraw.Draw(mask,'RGBA')
    draw.polygon(polygonpos,(0,0,0,255))
    del draw
    mask.save("mask.png")

# Get the Alpha band from the template
    tmplt = Image.open('mask.png')
    A = tmplt.split()[3]


#make one piece of triangle on transparent bg
    [R,G,B]=base.split()
    tri = Image.merge('RGBA', (R, G, B, A))

#crop it to the exact size of triangle!! to create primary pattern
#box (left, top , right, buttom)
    box =(originx,(originy-trih),(originx+triw),originy)
    pattern_plain=tri.crop(box)
    pattern_plain.save('pattern_plain.png')
    print('....pattern created....')


# add style to pattern
    pattern = pattern_plain
#pattern= pattern_plain.filter(ImageFilter.EDGE_ENHANCE)
    pattern.save('pattern_tri.png')
    print('....stylized pattern....')


#make square canvas for the output (wide = double size of height of primary pattern)
    canvas =Image.new('RGBA',(2*trih,2*trih), (255,255,255,0))
    canvas.save('tmpcanvas.png')
    pcanvas=Image.new('RGBA',(2*trih,2*trih), (255,255,255,0))




#put pattern on the canvas
#make sure to put the tip of the triangle at the center of the canvas
#because when we rotate the center of the object is the pivot point
#note: paste command require the coordinate of top left corner
#so point to paste the pattern is . . .
    ccenterx = int(trih-triw/2) 
    canvas.paste(pattern,(ccenterx,trih))


# start rotate the pattern around every 40 degree

    for i in range (0,360,40):
        tmpcanvas = canvas
        tmppat = canvas.rotate(i)
        canvas= Image.alpha_composite(tmpcanvas,tmppat)

# now we get half of the things
    half = canvas

#mirror the half and put in the space to create simple kaleidoscpoe effect
    mirror = im.mirror(half)
    half2= mirror.rotate(20)


#merge 2 half
    output=Image.alpha_composite(half,half2)


    output.save("final.png")

if __name__ == '__main__':
    makekaleido()


FINAL BOT

STEP 2 : moving Kaleido

The next step is to make the bot gives more feeling of Kaleidoscope. The nice part of Kaleidoscope is whhen one rotate it around and see the abtract patterns moving.

I try to make it the similar feeling here with the bot that each still image is so abstact but when one finish watching the whole loop of rotation then you will have clue about what is the source image.

Kaleidochart2.jpg

In the code, I made the variable 'n' and 'a' to be able to adjust the the result GIF file. n is the number of the rotation time a is the degree of the rotation

so they will affect how smooth the animated GIF going to be.


STEP 3 Making the bot works online

The @kaleidogif twiiter account is created to be the page for the bot. ( Before that I just did trial and errors with my own account)

Link : [KaleidoGIF twitter|https://twitter.com/kaleidogif]


Botscreen1.PNG

This is the example result.

Botexample1.png Kaleidogif1.gif

more examples of the final result !! Botexamplegif2.gif Kaleidoexample3.gif.gif Kaleidoexample2.gif


FINISHING STEP

Last step . . . for the summaery event , I improved a bit the quality of the GIF by editing the code to be able to make best quality it can and also the final GIF is not larger that the twitter status update limit.

So I added lines of code to get the size of the final GIF and resize it if it exceed the limit.


The complete code of the final version can be found here in the GitHub link. [1] feel free to download and develop it further if you are interested.  ;)