diff --git a/app/service/generate_image/service_generate_product_image.py b/app/service/generate_image/service_generate_product_image.py index 1c20a13..681e2b5 100644 --- a/app/service/generate_image/service_generate_product_image.py +++ b/app/service/generate_image/service_generate_product_image.py @@ -1,4 +1,196 @@ -#!/usr/bin/env python +# #!/usr/bin/env python +# # -*- coding: UTF-8 -*- +# """ +# @Project ：trinity_client +# @File ：service_att_recognition.py +# @Author ：周成融 +# @Date ：2023/7/26 12:01:05 +# @detail ： +# """ +# import json +# import logging +# import time +# +# import cv2 +# import numpy as np +# import redis +# import tritonclient.grpc as grpcclient +# from PIL import Image +# from tritonclient.utils import np_to_triton_dtype +# +# from app.core.config import * +# from app.schemas.generate_image import GenerateProductImageModel +# from app.service.generate_image.utils.upload_sd_image import upload_SDXL_image +# from app.service.utils.oss_client import oss_get_image +# +# logger = logging.getLogger() +# +# +# class GenerateProductImage: +# def __init__(self, request_data): +# if DEBUG is False: +# self.connection = pika.BlockingConnection(pika.ConnectionParameters(**RABBITMQ_PARAMS)) +# self.channel = self.connection.channel() +# # self.connection = pika.BlockingConnection(pika.ConnectionParameters(**RABBITMQ_PARAMS)) +# # self.channel = self.connection.channel() +# # self.minio_client = Minio(MINIO_URL, access_key=MINIO_ACCESS, secret_key=MINIO_SECRET, secure=MINIO_SECURE) +# self.grpc_client = grpcclient.InferenceServerClient(url=GPI_MODEL_URL) +# self.redis_client = redis.StrictRedis(host=REDIS_HOST, port=REDIS_PORT, db=REDIS_DB, decode_responses=True) +# self.category = "product_image" +# self.image_strength = request_data.image_strength +# self.batch_size = 1 +# self.product_type = request_data.product_type +# self.prompt = request_data.prompt +# self.image, self.image_size, self.left, self.top = pre_processing_image(request_data.image_url) +# self.tasks_id = request_data.tasks_id +# self.user_id = self.tasks_id[self.tasks_id.rfind('-') + 1:] +# self.gen_product_data = {'tasks_id': self.tasks_id, 'status': 'PENDING', 'message': "pending", 'image_url': ''} +# self.redis_client.set(self.tasks_id, json.dumps(self.gen_product_data)) +# self.redis_client.expire(self.tasks_id, 600) +# +# def callback(self, result, error): +# if error: +# self.gen_product_data['status'] = "FAILURE" +# self.gen_product_data['message'] = str(error) +# self.redis_client.set(self.tasks_id, json.dumps(self.gen_product_data)) +# else: +# # pil图像转成numpy数组 +# image = result.as_numpy("generated_inpaint_image") +# image_result = Image.fromarray(np.squeeze(image.astype(np.uint8))).resize(self.image_size) +# cropped_image = post_processing_image(image_result, self.left, self.top) +# image_url = upload_SDXL_image(cropped_image, user_id=self.user_id, category=f"{self.category}", file_name=f"{self.tasks_id}.png") +# self.gen_product_data['status'] = "SUCCESS" +# self.gen_product_data['message'] = "success" +# self.gen_product_data['image_url'] = str(image_url) +# self.redis_client.set(self.tasks_id, json.dumps(self.gen_product_data)) +# +# def read_tasks_status(self): +# status_data = self.redis_client.get(self.tasks_id) +# return json.loads(status_data), status_data +# +# def get_result(self): +# try: +# prompts = [self.prompt] * self.batch_size +# self.image = cv2.cvtColor(self.image, cv2.COLOR_BGR2RGB) +# self.image = cv2.resize(self.image, (1024, 1024)) +# images = [self.image.astype(np.uint8)] * self.batch_size +# +# if self.product_type == "single": +# text_obj = np.array(prompts, dtype="object").reshape(-1, 1) +# image_obj = np.array(images, dtype=np.uint8).reshape((-1, 1024, 1024, 3)) +# image_strength_obj = np.array(self.image_strength, dtype=np.float32).reshape(-1, 1) +# else: +# text_obj = np.array(prompts, dtype="object").reshape((1)) +# image_obj = np.array(images, dtype=np.uint8).reshape((1024, 1024, 3)) +# image_strength_obj = np.array(self.image_strength, dtype=np.float32).reshape((1)) +# +# # 假设 prompts、images 和 self.image_strength 已经定义 +# +# input_text = grpcclient.InferInput("prompt", text_obj.shape, np_to_triton_dtype(text_obj.dtype)) +# input_image = grpcclient.InferInput("input_image", image_obj.shape, "UINT8") +# input_image_strength = grpcclient.InferInput("image_strength", image_strength_obj.shape, np_to_triton_dtype(image_strength_obj.dtype)) +# +# input_text.set_data_from_numpy(text_obj) +# input_image.set_data_from_numpy(image_obj) +# input_image_strength.set_data_from_numpy(image_strength_obj) +# +# inputs = [input_text, input_image, input_image_strength] +# +# if self.product_type == "single": +# ctx = self.grpc_client.async_infer(model_name="stable_diffusion_xl_cnet_inpaint", inputs=inputs, callback=self.callback) +# else: +# ctx = self.grpc_client.async_infer(model_name=GPI_MODEL_NAME_OVERALL, inputs=inputs, callback=self.callback) +# +# time_out = 600 +# while time_out > 0: +# gen_product_data, _ = self.read_tasks_status() +# if gen_product_data['status'] in ["REVOKED", "FAILURE"]: +# ctx.cancel() +# break +# elif gen_product_data['status'] == "SUCCESS": +# break +# time_out -= 1 +# time.sleep(0.1) +# gen_product_data, _ = self.read_tasks_status() +# return gen_product_data +# except Exception as e: +# self.gen_product_data['status'] = "FAILURE" +# self.gen_product_data['message'] = str(e) +# self.redis_client.set(self.tasks_id, json.dumps(self.gen_product_data)) +# raise Exception(str(e)) +# finally: +# dict_gen_product_data, str_gen_product_data = self.read_tasks_status() +# if DEBUG is False: +# self.channel.basic_publish(exchange='', routing_key=GPI_RABBITMQ_QUEUES, body=str_gen_product_data) +# logger.info(f" [x] Sent to： {GPI_RABBITMQ_QUEUES} data：@@@@ {json.dumps(dict_gen_product_data, indent=4)}") +# +# +# def infer_cancel(tasks_id): +# redis_client = redis.StrictRedis(host=REDIS_HOST, port=REDIS_PORT, db=REDIS_DB, decode_responses=True) +# data = {'tasks_id': tasks_id, 'status': 'REVOKED', 'message': "revoked", 'data': 'revoked'} +# gen_product_data = json.dumps(data) +# redis_client.set(tasks_id, gen_product_data) +# return data +# +# +# def pre_processing_image(image_url): +# image = oss_get_image(bucket=image_url.split('/')[0], object_name=image_url[image_url.find('/') + 1:], data_type="PIL") +# # resize 原图至1024*1024 +# image = image.resize((int(1024 / image.height * image.width), 1024)) +# +# # 原始图片的尺寸 +# width, height = image.size +# +# new_height, new_width = 1024, 1024 +# # 创建一个新的画布，大小为添加 padding 后的尺寸，并设置为白色背景 +# pad_image = Image.new('RGBA', (new_width, new_height), (0, 0, 0, 0)) +# +# # 将原始图片粘贴到新的画布中心 +# left = (new_width - width) // 2 +# top = (new_height - height) // 2 +# pad_image.paste(image, (left, top)) +# +# # 将画布 resize 成宽度 1024，长度 1024 +# resized_image = pad_image.resize((1024, 1024)) +# image_size = (1024, 1024) +# +# if resized_image.mode in ('RGBA', 'LA') or (resized_image.mode == 'P' and 'transparency' in resized_image.info): +# # 创建白色背景 +# background = Image.new("RGB", image_size, (255, 255, 255)) +# # 将图片粘贴到白色背景上 +# background.paste(resized_image, mask=resized_image.split()[3]) +# image = np.array(background) +# image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB) +# return image, image_size, left, top +# +# +# def post_processing_image(image, left, top): +# resized_image = image.resize((int(image.width * (768 / image.height)), 768)) +# # 计算裁剪的坐标 +# left = (resized_image.width - 512) // 2 +# upper = 0 +# right = left + 512 +# lower = 768 +# +# # 进行裁剪 +# cropped_image = resized_image.crop((left, upper, right, lower)) +# return cropped_image +# +# +# if __name__ == '__main__': +# rd = GenerateProductImageModel( +# tasks_id="123-89", +# # prompt="", +# image_strength=0.7, +# prompt="The best quality, masterpiece,outwear, 8K realistic, HUD", +# image_url="aida-results/result_53381ada-ac64-11ef-ae9d-0242ac150002.png", +# product_type="overall" +# ) +# server = GenerateProductImage(rd) +# print(server.get_result()) + +# 旧版product +# !/usr/bin/env python # -*- coding: UTF-8 -*- """ @Project ：trinity_client @@ -34,14 +226,14 @@ class GenerateProductImage: # self.connection = pika.BlockingConnection(pika.ConnectionParameters(**RABBITMQ_PARAMS)) # self.channel = self.connection.channel() # self.minio_client = Minio(MINIO_URL, access_key=MINIO_ACCESS, secret_key=MINIO_SECRET, secure=MINIO_SECURE) - self.grpc_client = grpcclient.InferenceServerClient(url=GPI_MODEL_URL) + self.grpc_client = grpcclient.InferenceServerClient(url="10.1.1.243:18001") self.redis_client = redis.StrictRedis(host=REDIS_HOST, port=REDIS_PORT, db=REDIS_DB, decode_responses=True) self.category = "product_image" self.image_strength = request_data.image_strength self.batch_size = 1 self.product_type = request_data.product_type self.prompt = request_data.prompt - self.image, self.image_size, self.left, self.top = pre_processing_image(request_data.image_url) + self.image = pre_processing_image(request_data.image_url) self.tasks_id = request_data.tasks_id self.user_id = self.tasks_id[self.tasks_id.rfind('-') + 1:] self.gen_product_data = {'tasks_id': self.tasks_id, 'status': 'PENDING', 'message': "pending", 'image_url': ''} @@ -56,9 +248,9 @@ class GenerateProductImage: else: # pil图像转成numpy数组 image = result.as_numpy("generated_inpaint_image") - image_result = Image.fromarray(np.squeeze(image.astype(np.uint8))).resize(self.image_size) - cropped_image = post_processing_image(image_result, self.left, self.top) - image_url = upload_SDXL_image(cropped_image, user_id=self.user_id, category=f"{self.category}", file_name=f"{self.tasks_id}.png") + image_result = Image.fromarray(np.squeeze(image.astype(np.uint8))) + # cropped_image = post_processing_image(image_result, self.left, self.top) + image_url = upload_SDXL_image(image_result, user_id=self.user_id, category=f"{self.category}", file_name=f"{self.tasks_id}.png") self.gen_product_data['status'] = "SUCCESS" self.gen_product_data['message'] = "success" self.gen_product_data['image_url'] = str(image_url) @@ -72,7 +264,7 @@ class GenerateProductImage: try: prompts = [self.prompt] * self.batch_size self.image = cv2.cvtColor(self.image, cv2.COLOR_BGR2RGB) - self.image = cv2.resize(self.image, (1024, 1024)) + # self.image = cv2.resize(self.image, (1024, 1024)) images = [self.image.astype(np.uint8)] * self.batch_size if self.product_type == "single": @@ -81,7 +273,7 @@ class GenerateProductImage: image_strength_obj = np.array(self.image_strength, dtype=np.float32).reshape(-1, 1) else: text_obj = np.array(prompts, dtype="object").reshape((1)) - image_obj = np.array(images, dtype=np.uint8).reshape((1024, 1024, 3)) + image_obj = np.array(images, dtype=np.uint8).reshape((768, 512, 3)) image_strength_obj = np.array(self.image_strength, dtype=np.float32).reshape((1)) # 假设 prompts、images 和 self.image_strength 已经定义 @@ -99,7 +291,7 @@ class GenerateProductImage: if self.product_type == "single": ctx = self.grpc_client.async_infer(model_name="stable_diffusion_xl_cnet_inpaint", inputs=inputs, callback=self.callback) else: - ctx = self.grpc_client.async_infer(model_name=GPI_MODEL_NAME_OVERALL, inputs=inputs, callback=self.callback) + ctx = self.grpc_client.async_infer(model_name="diffusion_ensemble_all", inputs=inputs, callback=self.callback) time_out = 600 while time_out > 0: @@ -135,33 +327,25 @@ def infer_cancel(tasks_id): def pre_processing_image(image_url): image = oss_get_image(bucket=image_url.split('/')[0], object_name=image_url[image_url.find('/') + 1:], data_type="PIL") - # resize 原图至1024*1024 - image = image.resize((int(1024 / image.height * image.width), 1024)) - - # 原始图片的尺寸 + # 调整图片高度为768像素，保持宽高比 width, height = image.size + new_height = 768 + new_width = int(width * (new_height / height)) + resized_image = image.resize((new_width, new_height)) - new_height, new_width = 1024, 1024 - # 创建一个新的画布，大小为添加 padding 后的尺寸，并设置为白色背景 - pad_image = Image.new('RGBA', (new_width, new_height), (0, 0, 0, 0)) + # 创建一个512x768的透明图片 + result_image = Image.new("RGBA", (512, 768), (0, 0, 0, 0)) - # 将原始图片粘贴到新的画布中心 - left = (new_width - width) // 2 - top = (new_height - height) // 2 - pad_image.paste(image, (left, top)) + # 计算需要粘贴的位置，使图片居中 + x_offset = (512 - new_width) // 2 + y_offset = 0 - # 将画布 resize 成宽度 1024，长度 1024 - resized_image = pad_image.resize((1024, 1024)) - image_size = (1024, 1024) + # 将调整大小后的图片粘贴到透明图片上 + result_image.paste(resized_image, (x_offset, y_offset)) - if resized_image.mode in ('RGBA', 'LA') or (resized_image.mode == 'P' and 'transparency' in resized_image.info): - # 创建白色背景 - background = Image.new("RGB", image_size, (255, 255, 255)) - # 将图片粘贴到白色背景上 - background.paste(resized_image, mask=resized_image.split()[3]) - image = np.array(background) - image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB) - return image, image_size, left, top + image = np.array(result_image) + image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB) + return image def post_processing_image(image, left, top):