Rating: 4.0

> Qubic Rube
>
> 300 points
>
> Please continue to solve Rubic's Cube and read QR code.
>
> http://qubicrube.pwn.seccon.jp:33654/

One of the reasons I love CTFs is because they force me to learn new technologies quickly. This challenge was awesome because I learned image manipulation and QR code scanning in Python - something I wanted to try for a while.

When we open the referenced site we are shown a spinning Rubics Cube with sides containing QR codes:

![image1.png](https://0xd13a.github.io/ctfs/seccon2017/qubic-rube/image1.png)

The images can be loaded directly and analyzed. As we [decode](https://zxing.org/w/decode.jspx) the codes we see that one of them contains a reference to the next page in sequence (50 of them total):

![image2.png](https://0xd13a.github.io/ctfs/seccon2017/qubic-rube/image2.png)

However, as we progress from page to page the images get more and more mangled:

![image3.png](https://0xd13a.github.io/ctfs/seccon2017/qubic-rube/image3.png)

It's clear that we have to reconstruct proper sides from separate pieces. Here is the sequence of steps that we have to follow:

- Load all images on a page
- Split each in 9 pieces (3x3)
- For each piece determine the color side it belongs to
- "Normalize" pieces by rotating corner pieces to be in the top-left position, and sides to be in the top-middle position
- Join pieces for each side in all possible combinations to see if we can extract the text
- If the text is a flag - show it
- If the text is the reference to the next page - follow it and start again

Let's put this algorithm into a script:

```python
import qrtools
import urllib2
from PIL import Image
import os

SIZE = 82
TEST_IMG = "test.png"

color_map = {}

color_map[(196, 30, 58)] = 0
color_map[(255, 88, 0)] = 1
color_map[(255, 255, 255)] = 2
color_map[(0, 81, 186)] = 3
color_map[(0, 158, 96)] = 4
color_map[(255, 213, 0)] = 5

permutations = [[1,2,3,4], [1,2,4,3], [1,3,2,4], [1,3,4,2], [1,4,2,3], [1,4,3,2],
[2,1,3,4], [2,1,4,3], [2,3,1,4], [2,3,4,1], [2,4,1,3], [2,4,3,1],
[3,1,2,4], [3,1,4,2], [3,2,1,4], [3,2,4,1], [3,4,1,2], [3,4,2,1],
[4,1,2,3], [4,1,3,2], [4,2,1,3], [4,2,3,1], [4,3,1,2], [4,3,2,1]
]

def process(file, center, corners, sides):
orig_img = Image.open(file)

# cut out 9 pieces
for x in range(3):
for y in range(3):
img = orig_img.copy()
img = img.crop((x*SIZE, y*SIZE, x*SIZE+SIZE, y*SIZE+SIZE))
colors = img.getcolors(256) #put a higher value if there are many colors in your image

# determine piece color
for c in colors:
if c[1] in color_map:
colorid = color_map[c[1]]

# normalize by rotation
if x == 0 and y > 0: img = img.rotate(270)
if x == 2 and y < 2: img = img.rotate(90)
if x > 0 and y == 2: img = img.rotate(180)

# store piece in proper bucket
if x == 1 and y == 1: center[colorid] = img
else:
if x == 1 or y == 1: sides[colorid].append(img)
else: corners[colorid].append(img)

def combine(center, corners, sides):
global permutations

# recombine pieces all possible ways
for colorid in range(6):
for corn in range(len(permutations)):
for side in range(len(permutations)):
img = Image.new("RGB",(SIZE*3,SIZE*3))

img.paste(center[colorid],(SIZE,SIZE))

# paste and rotate corners
for x in range(4):
pos = permutations[corn][x]
if pos == 1: img.paste(corners[colorid][x],(0,0))
if pos == 2: img.paste(corners[colorid][x].rotate(-90),(SIZE*2,0))
if pos == 3: img.paste(corners[colorid][x].rotate(-180),(SIZE*2,SIZE*2))
if pos == 4: img.paste(corners[colorid][x].rotate(-270),(0,SIZE*2))

# paste and rotate sides
for x in range(4):
pos = permutations[side][x]
if pos == 1: img.paste(sides[colorid][x],(SIZE,0))
if pos == 2: img.paste(sides[colorid][x].rotate(-90),(SIZE*2,SIZE))
if pos == 3: img.paste(sides[colorid][x].rotate(-180),(SIZE,SIZE*2))
if pos == 4: img.paste(sides[colorid][x].rotate(-270),(0,SIZE))

img.save(TEST_IMG)

qr = qrtools.QR()
qr.decode(TEST_IMG)

os.remove(TEST_IMG)

# see if we found the link to the next page
if "seccon.jp" in qr.data:
return qr.data[qr.data.rfind("/")+1:]

# print the flag if found
if "SECCON" in qr.data:
print qr.data
return None

# starting image prefix
pref = "01000000000000000000"

while True:
print "---"

corners = []
sides = []
center = []

for x in range(6):
corners.append([])
sides.append([])
center.append(None)

# download all sides
for x in "RLUDFB":
file = "%s_%s.png" % (pref,x)

open(file,"wb").write(urllib2.urlopen('http://qubicrube.pwn.seccon.jp:33654/images/' + file).read())
process(file, center, corners, sides)

# find the next link or the flag
pref = combine(center, corners, sides)

if pref == None:
print "Not found"
break
else:
print "Found " + pref
```

After running for a while the script gets us the answer:

```
...
Found 4882153b757d0af86d97
---
Found 49d06dfaeaefaa612e72
---
SECCON 2017 Online CTF
SECCON 2017 Online CTF
SECCON 2017 Online CTF
SECCON 2017 Online CTF
SECCON 2017 Online CTF
Found 504ded069e4db4e3bef9
---
SECCON{Thanks to Denso Wave for inventing the QR code}
SECCON{Thanks to Denso Wave for inventing the QR code}
SECCON{Thanks to Denso Wave for inventing the QR code}
SECCON{Thanks to Denso Wave for inventing the QR code}
SECCON 2017 Online CTF
SECCON 2017 Online CTF
SECCON 2017 Online CTF
SECCON 2017 Online CTF
SECCON 2017 Online CTF
Not found
```

The flag is ```SECCON{Thanks to Denso Wave for inventing the QR code}```.

Original writeup (https://0xd13a.github.io/ctfs/seccon2017/qubic-rube/).