feat:

fix:  替换项目中所有mmcv的依赖

app/service/attribute/service_att_recognition.py

@@ -3,7 +3,6 @@
 from pprint import pprint
 
 import cv2
-import mmcv
 import numpy as np
 import pandas as pd
 import torch
@@ -12,6 +11,7 @@ from minio import Minio
 
 from app.core.config import settings, DESIGN_MODEL_URL
 from app.schemas.attribute_retrieve import AttributeRecognitionModel
+from app.service.utils.image_normalize import my_imnormalize
 from app.service.utils.new_oss_client import oss_get_image
 
 minio_client = Minio(settings.MINIO_URL, access_key=settings.MINIO_ACCESS, secret_key=settings.MINIO_SECRET, secure=settings.MINIO_SECURE)
@@ -109,10 +109,9 @@ class AttributeRecognition:
 
     @staticmethod
     def preprocess(img):
-        img = mmcv.imread(img)
         img_scale = (224, 224)
         img = cv2.resize(img, img_scale)
-        img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+        img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
         preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
         return preprocessed_img
 

app/service/attribute/service_category_recognition.py

@@ -10,7 +10,6 @@
 from minio import Minio
 from skimage import transform
 import cv2
-import mmcv
 import numpy as np
 import pandas as pd
 import tritonclient.http as httpclient
@@ -18,6 +17,7 @@ import torch
 
 from app.core.config import settings, DESIGN_MODEL_URL
 from app.schemas.attribute_retrieve import CategoryRecognitionModel
+from app.service.utils.image_normalize import my_imnormalize
 from app.service.utils.new_oss_client import oss_get_image
 
 minio_client = Minio(settings.MINIO_URL, access_key=settings.MINIO_ACCESS, secret_key=settings.MINIO_SECRET, secure=settings.MINIO_SECURE)
@@ -39,11 +39,10 @@ class CategoryRecognition:
 
     @staticmethod
     def preprocess(img):
-        img = mmcv.imread(img)
         # ori_shape = img.shape[:2]
         img_scale = (224, 224)
         img = cv2.resize(img, img_scale)
-        img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+        img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
         preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
         return preprocessed_img
 

app/service/brand_dna/service.py

@@ -1,7 +1,6 @@
 import logging
 
 import cv2
-import mmcv
 import numpy as np
 import pandas as pd
 import torch
@@ -14,6 +13,7 @@ from app.core.config import settings
 from app.schemas.brand_dna import BrandDnaModel
 from app.service.attribute.config import const
 from app.service.utils.generate_uuid import generate_uuid
+from app.service.utils.image_normalize import my_imnormalize
 from app.service.utils.new_oss_client import oss_upload_image, oss_get_image
 
 minio_client = Minio(settings.MINIO_URL, access_key=settings.MINIO_ACCESS, secret_key=settings.MINIO_SECRET, secure=settings.MINIO_SECURE)
@@ -202,7 +202,7 @@ class BrandDna:
     # 服装分割预处理
     @staticmethod
     def seg_product_preprocess(image):
-        img = mmcv.imread(image)
+        img = image
         ori_shape = img.shape[:2]
         img_scale_w, img_scale_h = ori_shape
         if ori_shape[0] > 1024:
@@ -211,9 +211,9 @@ class BrandDna:
             img_scale_h = 1024
         # 如果图片size任意一边 大于 1024， 则会resize 成1024
         if ori_shape != (img_scale_w, img_scale_h):
-            # mmcv.imresize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
+            # my_imnormalize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
             img = cv2.resize(img, (img_scale_h, img_scale_w))
-        img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+        img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
         preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
         return preprocessed_img, ori_shape
 
@@ -227,11 +227,10 @@ class BrandDna:
     # 类别检测模型预处理
     @staticmethod
     def category_preprocess(img):
-        img = mmcv.imread(img)
         # ori_shape = img.shape[:2]
         img_scale = (224, 224)
         img = cv2.resize(img, img_scale)
-        img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+        img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
         preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
         return preprocessed_img
 

app/service/design_batch/utils/design_ensemble.py

@@ -10,13 +10,13 @@
 import logging
 
 import cv2
-import mmcv
 import numpy as np
 import torch
 import torch.nn.functional as F
 import tritonclient.http as httpclient
 
 from app.core.config import DESIGN_MODEL_URL, DESIGN_MODEL_NAME
+from app.service.utils.image_normalize import my_imnormalize
 
 """
     keypoint
@@ -25,13 +25,13 @@ from app.core.config import DESIGN_MODEL_URL, DESIGN_MODEL_NAME
 
 
 def keypoint_preprocess(img_path):
-    img = mmcv.imread(img_path)
+    img = img_path
     img_scale = (256, 256)
     h, w = img.shape[:2]
     img = cv2.resize(img, img_scale)
     w_scale = img_scale[0] / w
     h_scale = img_scale[1] / h
-    img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+    img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
     preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
     return preprocessed_img, (w_scale, h_scale)
 
@@ -74,7 +74,7 @@ def keypoint_postprocess(output, scale_factor):
 
 # KNet
 def seg_preprocess(img_path):
-    img = mmcv.imread(img_path)
+    img = img_path
     ori_shape = img.shape[:2]
     img_scale_w, img_scale_h = ori_shape
     if ori_shape[0] > 1024:
@@ -83,9 +83,9 @@ def seg_preprocess(img_path):
         img_scale_h = 1024
     # 如果图片size任意一边 大于 1024， 则会resize 成1024
     if ori_shape != (img_scale_w, img_scale_h):
-        # mmcv.imresize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
+        # my_imnormalize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
         img = cv2.resize(img, (img_scale_h, img_scale_w))
-    img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+    img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
     preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
     return preprocessed_img, ori_shape
 

app/service/design_fast/utils/design_ensemble.py

@@ -10,12 +10,12 @@
 import logging
 
 import cv2
-import mmcv
 import numpy as np
 import torch
 import tritonclient.http as httpclient
 
 from app.core.config import DESIGN_MODEL_URL, DESIGN_MODEL_NAME
+from app.service.utils.image_normalize import my_imnormalize
 
 """
     keypoint
@@ -24,14 +24,14 @@ from app.core.config import DESIGN_MODEL_URL, DESIGN_MODEL_NAME
 
 
 def keypoint_preprocess(img_path):
-    img = mmcv.imread(img_path)
+    img = img_path
     img = cv2.copyMakeBorder(img, 25, 25, 25, 25, cv2.BORDER_CONSTANT, value=[255, 255, 255])
     img_scale = (256, 256)
     h, w = img.shape[:2]
     img = cv2.resize(img, img_scale)
     w_scale = img_scale[0] / w
     h_scale = img_scale[1] / h
-    img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+    img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
     preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
     return preprocessed_img, (w_scale, h_scale)
 
@@ -78,7 +78,7 @@ def keypoint_postprocess(output, scale_factor):
 
 # KNet
 def seg_preprocess(img_path):
-    img = mmcv.imread(img_path)
+    img = img_path
     ori_shape = img.shape[:2]
     img_scale_w, img_scale_h = ori_shape
     if ori_shape[0] > 1024:
@@ -87,12 +87,12 @@ def seg_preprocess(img_path):
         img_scale_h = 1024
     # 如果图片size任意一边 大于 1024， 则会resize 成1024
     if ori_shape != (img_scale_w, img_scale_h):
-        # mmcv.imresize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
+        # my_imnormalize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
         img = cv2.resize(img, (img_scale_h, img_scale_w))
 
     # 扩充25的白边
     img = cv2.copyMakeBorder(img, 25, 25, 25, 25, cv2.BORDER_CONSTANT, value=[255, 255, 255])
-    img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+    img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
     preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
     return preprocessed_img, ori_shape
 

app/service/generate_image/service_agent_tool_generate_image.py

@@ -11,7 +11,6 @@ import logging
 import uuid
 
 import cv2
-import mmcv
 import numpy as np
 import pandas as pd
 import torch
@@ -21,6 +20,7 @@ from minio import Minio
 from tritonclient.utils import np_to_triton_dtype
 
 from app.core.config import settings, FAST_GI_MODEL_URL, GI_MODEL_URL, DESIGN_MODEL_URL, FAST_GI_MODEL_NAME, GI_MODEL_NAME
+from app.service.utils.image_normalize import my_imnormalize
 from app.service.utils.new_oss_client import oss_upload_image
 
 logger = logging.getLogger()
@@ -86,10 +86,9 @@ class AgentToolGenerateImage:
 
     @staticmethod
     def preprocess(img):
-        img = mmcv.imread(img)
         img_scale = (224, 224)
         img = cv2.resize(img, img_scale)
-        img = mmcv.imnormalize(
+        img = my_imnormalize(
             img,
             mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]),
             to_rgb=True)

app/service/generate_image/service_generate_image.py

@@ -189,10 +189,10 @@ if __name__ == '__main__':
         tasks_id="123-89",
         prompt="a single item of sketch of dress, 4k, white background",
         image_url="aida-collection-element/89/Sketchboard/95f20cdc-e059-435c-b8b1-d04cc9e80c3d.png",
-        mode='img2img',
+        mode='txt2img',
         category="sketch",
         gender="Female",
-        version="fast"
+        version="hight"
     )
     server = GenerateImage(rd)
     print(server.get_result())

app/service/generate_image/utils/image_processing.py

@@ -2,23 +2,23 @@ import logging
 import time
 
 import cv2
-import mmcv
 import numpy as np
 import torch
 import tritonclient.http as httpclient
 
 from app.core.config import settings, DESIGN_MODEL_URL, DESIGN_MODEL_NAME
 from app.service.generate_image.utils.upload_sd_image import upload_stain_png_sd, upload_face_png_sd
+from app.service.utils.image_normalize import my_imnormalize
 
 logger = logging.getLogger()
 
 
 def seg_preprocess(img_path):
-    img = mmcv.imread(img_path)
+    img = img_path
     ori_shape = img.shape[:2]
     img_scale = ori_shape
     img = cv2.resize(img, img_scale)
-    img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+    img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
     preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
     return preprocessed_img, ori_shape
 
@@ -242,10 +242,9 @@ def stain_detection(image, user_id, category, tasks_id, spot_size=100):
 
 def generate_category_recognition(image, gender):
     def preprocess(img):
-        img = mmcv.imread(img)
         img_scale = (224, 224)
         img = cv2.resize(img, img_scale)
-        img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+        img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
         preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
         return preprocessed_img
 

app/service/lineart/service.py

@@ -1,7 +1,6 @@
 import logging
 
 import cv2
-import mmcv
 import numpy as np
 import torch
 import torch.nn.functional as F
@@ -10,6 +9,7 @@ from minio import Minio
 from app.core.config import settings
 from app.core.config import DESIGN_MODEL_URL
 from app.schemas.image2sketch import Image2SketchModel
+from app.service.utils.image_normalize import my_imnormalize
 from app.service.utils.new_oss_client import oss_get_image, oss_upload_image
 
 logger = logging.getLogger()
@@ -67,7 +67,7 @@ class LineArtService:
 
     @staticmethod
     def line_art_preprocess(image):
-        img = mmcv.imread(image)
+        img = image
         ori_shape = img.shape[:2]
         img_scale_w, img_scale_h = ori_shape
         if ori_shape[0] > 1024:
@@ -76,9 +76,9 @@ class LineArtService:
             img_scale_h = 1024
         # 如果图片size任意一边 大于 1024， 则会resize 成1024
         if ori_shape != (img_scale_w, img_scale_h):
-            # mmcv.imresize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
+            # my_imnormalize(img, img_scale_h, img_scale_w) # 老代码 引以为戒！哈哈哈~ h和w写反了
             img = cv2.resize(img, (img_scale_h, img_scale_w))
-        img = mmcv.imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
+        img = my_imnormalize(img, mean=np.array([123.675, 116.28, 103.53]), std=np.array([58.395, 57.12, 57.375]), to_rgb=True)
         preprocessed_img = np.expand_dims(img.transpose(2, 0, 1), axis=0)
         return preprocessed_img, ori_shape
 

app/service/utils/image_normalize.py

@@ -0,0 +1,27 @@
+import cv2
+import np
+
+
+def my_imnormalize(img, mean, std, to_rgb=True):
+    """Inplace normalize an image with mean and std.
+
+    Args:
+        img (ndarray): Image to be normalized.
+        mean (ndarray): The mean to be used for normalize.
+        std (ndarray): The std to be used for normalize.
+        to_rgb (bool): Whether to convert to rgb.
+
+    Returns:
+        ndarray: The normalized image.
+    """
+    # cv2 inplace normalization does not accept uint8
+    img = img.copy().astype(np.float32)
+
+    assert img.dtype != np.uint8
+    mean = np.float64(mean.reshape(1, -1))
+    stdinv = 1 / np.float64(std.reshape(1, -1))
+    if to_rgb:
+        cv2.cvtColor(img, cv2.COLOR_BGR2RGB, img)  # inplace
+    cv2.subtract(img, mean, img)  # inplace
+    cv2.multiply(img, stdinv, img)  # inplace
+    return img