52 lines
1.8 KiB
Python
52 lines
1.8 KiB
Python
import os
|
|
import torch
|
|
import numpy as np
|
|
from PIL import Image, ImageOps, ImageSequence
|
|
|
|
from ..client_s3 import get_s3_instance
|
|
S3_INSTANCE = get_s3_instance()
|
|
|
|
|
|
class LoadImageS3:
|
|
@classmethod
|
|
def INPUT_TYPES(s):
|
|
input_dir = os.getenv("S3_INPUT_DIR")
|
|
files = S3_INSTANCE.get_files(prefix=input_dir)
|
|
return {"required":
|
|
{"image": (sorted(files), {"image_upload": False})},
|
|
}
|
|
|
|
CATEGORY = "ComfyS3"
|
|
RETURN_TYPES = ("IMAGE", "MASK")
|
|
FUNCTION = "load_image"
|
|
|
|
def load_image(self, image):
|
|
s3_path = os.path.join(os.getenv("S3_INPUT_DIR"), image)
|
|
image_path = S3_INSTANCE.download_file(s3_path=s3_path, local_path=f"input/{image}")
|
|
|
|
img = Image.open(image_path)
|
|
output_images = []
|
|
output_masks = []
|
|
for i in ImageSequence.Iterator(img):
|
|
i = ImageOps.exif_transpose(i)
|
|
if i.mode == 'I':
|
|
i = i.point(lambda i: i * (1 / 255))
|
|
image = i.convert("RGB")
|
|
image = np.array(image).astype(np.float32) / 255.0
|
|
image = torch.from_numpy(image)[None,]
|
|
if 'A' in i.getbands():
|
|
mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0
|
|
mask = 1. - torch.from_numpy(mask)
|
|
else:
|
|
mask = torch.zeros((64,64), dtype=torch.float32, device="cpu")
|
|
output_images.append(image)
|
|
output_masks.append(mask.unsqueeze(0))
|
|
|
|
if len(output_images) > 1:
|
|
output_image = torch.cat(output_images, dim=0)
|
|
output_mask = torch.cat(output_masks, dim=0)
|
|
else:
|
|
output_image = output_images[0]
|
|
output_mask = output_masks[0]
|
|
|
|
return (output_image, output_mask) |