Support HEIC format (iPhone snapshot) and image preprocessing

app.py

@@ -24,7 +24,7 @@ from preprocess.humanparsing.run_parsing import Parsing
 from preprocess.openpose.run_openpose import OpenPose
 from detectron2.data.detection_utils import convert_PIL_to_numpy,_apply_exif_orientation
 from torchvision.transforms.functional import to_pil_image
-
+import pillow_heif  # HEIC 이미지 처리용 (아이폰 촬영 사진 포맷)
 
 def pil_to_binary_mask(pil_image, threshold=0):
     np_image = np.array(pil_image)
@@ -121,6 +121,41 @@ pipe = TryonPipeline.from_pretrained(
 )
 pipe.unet_encoder = UNet_Encoder
 
+
+# 이미지 전처리 함수
+def preprocess_image(image):
+    # HEIC 이미지 처리
+    if isinstance(image, np.ndarray):
+        image = Image.fromarray(image)
+    
+    # HEIC 이미지를 JPEG로 변환
+    try:
+        output = io.BytesIO()
+        image.convert("RGB").save(output, format="JPEG")
+        image = Image.open(output)
+    except Exception as e:
+        print(f"Error converting image: {e}")
+
+    # 이미지 크기 가져오기
+    width, height = image.size
+    
+    # 3:4 비율로 중앙 자르기
+    target_width = int(min(width, height * (3 / 4)))
+    target_height = int(min(height, width * (4 / 3)))
+    left = (width - target_width) / 2
+    top = (height - target_height) / 2
+    right = (width + target_width) / 2
+    bottom = (height + target_height) / 2
+    
+    # 이미지 자르기
+    cropped_img = image.crop((left, top, right, bottom))
+    
+    # 768x1024로 리사이징
+    resized_img = cropped_img.resize((768, 1024), resample=Image.Resampling.LANCZOS)
+    
+    return resized_img
+
+
 @spaces.GPU
 def start_tryon(dict,garm_img,garment_des,is_checked,denoise_steps,seed, is_checked_crop):
     device = "cuda"
@@ -133,15 +168,15 @@ def start_tryon(dict,garm_img,garment_des,is_checked,denoise_steps,seed, is_chec
     human_img_orig = dict["background"].convert("RGB")    
     
     if is_checked_crop:
-        width, height = human_img_orig.size
-        target_width = int(min(width, height * (3 / 4)))
-        target_height = int(min(height, width * (4 / 3)))
-        left = (width - target_width) / 2
-        top = (height - target_height) / 2
-        right = (width + target_width) / 2
-        bottom = (height + target_height) / 2
-        cropped_img = human_img_orig.crop((left, top, right, bottom))
-        crop_size = cropped_img.size
+        # width, height = human_img_orig.size
+        # target_width = int(min(width, height * (3 / 4)))
+        # target_height = int(min(height, width * (4 / 3)))
+        # left = (width - target_width) / 2
+        # top = (height - target_height) / 2
+        # right = (width + target_width) / 2
+        # bottom = (height + target_height) / 2
+        # cropped_img = human_img_orig.crop((left, top, right, bottom))
+        # crop_size = cropped_img.size
         human_img = cropped_img.resize((768,1024))
     else:
         human_img = human_img_orig.resize((768,1024))
@@ -301,6 +336,14 @@ with image_blocks as demo:
         #         seed = gr.Number(label="Seed", minimum=-1, maximum=2147483647, step=1, value=42)
 
     # is_checked = gr.Number(value=True)
+
+    # 이미지 업로드 시 전처리
+    imgs.upload(
+        fn=preprocess_image,
+        inputs=imgs,
+        outputs=imgs,  # 전처리된 이미지를 ImageEditor에 다시 표시
+    )
+
     is_checked_crop = True
     denoise_steps = 30
     seed = 42

requirements.txt

@@ -21,4 +21,5 @@ fvcore
 cloudpickle
 omegaconf
 pycocotools
-huggingface_hub==0.25.0
+huggingface_hub==0.25.0
+pillow-heif