File size: 1,484 Bytes
742cbff
86c3cf3
742cbff
86c3cf3
b8adf41
86c3cf3
 
b8adf41
86c3cf3
b8adf41
 
 
86c3cf3
b8adf41
86c3cf3
 
 
b8adf41
86c3cf3
 
 
b8adf41
86c3cf3
 
 
 
 
 
742cbff
86c3cf3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b8adf41
 
86c3cf3
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
import gradio as gr
from PIL import Image

def pixel_block_downsample(img, block_size):
    """
    Downsample the image by shrinking it and then re-upscaling
    with nearest neighbor to create large, chunky 'pixel blocks'.
    """
    # Ensure the image is in RGB
    if img.mode != "RGB":
        img = img.convert("RGB")
    
    w, h = img.size
    
    # Avoid zero or negative
    if block_size < 1:
        block_size = 1
    
    # Compute the size to downscale to
    new_w = max(1, w // block_size)
    new_h = max(1, h // block_size)
    
    # Downscale using a high-quality filter (e.g. BICUBIC or LANCZOS)
    img_small = img.resize((new_w, new_h), resample=Image.BICUBIC)
    
    # Upscale back using nearest-neighbor for the chunky pixel look
    img_blocky = img_small.resize((w, h), resample=Image.NEAREST)
    return img_blocky

def gradio_app():
    # Gradio interface with a slider for the "Pixel Block Size"
    iface = gr.Interface(
        fn=pixel_block_downsample,
        inputs=[
            gr.Image(type="pil", label="Upload Image"),
            gr.Slider(
                minimum=1, maximum=64, step=1, value=8,
                label="Choose Pixel Block Size"
            )
        ],
        outputs=gr.Image(type="pil", label="Downsampled Output"),
        title="Pixel Block Downsampler"
    )
    return iface

if __name__ == "__main__":
    # For local testing; on Hugging Face Spaces, this script is run automatically.
    gradio_app().launch()