Tags: proccessing image
Rating:
# Satan's Jigsaw
## Description
> Oh no! I dropped my pixels on the floor and they're all muddled up! It's going to take me years to sort all 90,000 of these again :(
Attached is a huge zip file which take ~ 30 min to decompress...
## Solution
In the zip file, there are 90,000 images named `n.jpg` with `n` being a random number. Every image contains only 1 pixel. By opening a few of them, we see that they have different colors. As the title hints for a jigsaw, let's reconstruct a 300x300 image with all those pixels in the order given by their file name.
```python
from PIL import Image
import os
import numpy as np
docs = os.listdir("jigsaw/chall")
doc_ints = [int(x.split(".")[0]) for x in docs]
doc_ints.sort()
i = 0
j = 0
image = [[None for k in range(300)] for l in range(300)]
for d in doc_ints:
image[i][j] = Image.open("jigsaw/chall/{}.jpg".format(d)).load()[0,0]
j += 1
if j == 300:
i += 1
j = 0
image = np.array(image, dtype=np.uint8)
new_image = Image.fromarray(image)
new_image.save("jigsaw.png")
```
Yeah we get back an image that makes sense!
We scan the QR code and get back our flag.
Flag: `rtcp{d1d-you_d0_7his_by_h4nd?}`
