Imageai用flask多线程部署出错

最新推荐文章于 2022-04-16 13:22:49 发布

weixin_43412231

最新推荐文章于 2022-04-16 13:22:49 发布

阅读量1.3k

点赞数 1

分类专栏： imageai python flask 文章标签： tensorflow debug flask

本文链接：https://blog.csdn.net/weixin_43412231/article/details/103715857

版权

python 同时被 3 个专栏收录

6 篇文章 0 订阅

订阅专栏

imageai

1 篇文章 0 订阅

订阅专栏

flask

1 篇文章 0 订阅

订阅专栏

环境简要说明

python3.7
设计主要相关依赖包版本

tensorflow = 1.15.0
keras = 2.3.1
flask = 1.1.1
imageai = 2.1.5

文件说明

init.py

from imageai.Detection import ObjectDetection
from flask import Flask
import tensorflow as tf
# 省略
detector = ObjectDetection()
detector.setModelTypeAsRetinaNet()
detector.setModelPath("./people/app/image_recognition/resnet50_coco_best_v2.0.1.h5")
detector.loadModel(detection_speed='fast')
# 省略
app = Flask(__name__)
app.config.from_object("food.config")

views.py

# 省略其他库的导入
from .. import app, detector

def cutting_image(filename):
    custom_objects = detector.CustomObjects(fork=True)
    output_objects_array, detections, detected_objects_image_array = detector.detectCustomObjectsFromImage(input_image=filename, 
        output_type = 'array',
        minimum_percentage_probability=27,
        extract_detected_objects = True,
        custom_objects=custom_objects)

@xxx.route('/xxxx',methods=['POST'])
def get_msg():
	# 省略
	msg = cutting_image(filename)
	# 省略
	return result

运行代码

运行flask程序，设置为多线程

python run.py runservere -h 0.0.0.0 -p 5000 --thread

这里没什么问题

现在开始调用接口

r = requests.post("http://127.0.0.1:5000/xxxx", data=body)

返回结果如下

2019-12-26 17:17:29,911:ERROR:food.app:Exception on /recommend/recognition_food [POST]
Traceback (most recent call last):
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/imageai/Detection/__init__.py", line 717, in detectCustomObjectsFromImage
    _, _, detections = model.predict_on_batch(np.expand_dims(image, axis=0))
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/keras/engine/training.py", line 1505, in predict_on_batch
    self._make_predict_function()
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/keras/engine/training.py", line 557, in _make_predict_function
    **kwargs)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/keras/backend/tensorflow_backend.py", line 3010, in function
    return Function(inputs, outputs, updates=updates, **kwargs)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/keras/backend/tensorflow_backend.py", line 2808, in __init__
    with tf.control_dependencies(self.outputs):
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 5254, in control_dependencies
    return get_default_graph().control_dependencies(control_inputs)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 4688, in control_dependencies
    c = self.as_graph_element(c)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 3607, in as_graph_element
    return self._as_graph_element_locked(obj, allow_tensor, allow_operation)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py", line 3686, in _as_graph_element_locked
    raise ValueError("Tensor %s is not an element of this graph." % obj)
ValueError: Tensor Tensor("regression/concat:0", shape=(?, ?, 4), dtype=float32) is not an element of this graph.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 2446, in wsgi_app
    response = self.full_dispatch_request()
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1951, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1820, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/_compat.py", line 39, in reraise
    raise value
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1949, in full_dispatch_request
    rv = self.dispatch_request()
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1935, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/home/yolanda/MyYolanda/user_analyze/food_recommend/project/food/app/image_recognition/views.py", line 198, in get_fod_msg
    all_foods_msg = recognition_all_food(path_open=image_path)
  File "/home/yolanda/MyYolanda/user_analyze/food_recommend/project/food/app/image_recognition/views.py", line 140, in recognition_all_food
    images_base64 = cutting_image(path_open)
  File "/home/yolanda/MyYolanda/user_analyze/food_recommend/project/food/app/image_recognition/views.py", line 45, in cutting_image
    custom_objects=custom_objects)           # 指定所要检测的物品类型，默认为全部
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/imageai/Detection/__init__.py", line 920, in detectCustomObjectsFromImage
    "Ensure you specified correct input image, input type, output type and/or output image path ")
ValueError: Ensure you specified correct input image, input type, output type and/or output image path

之前遇到过这问题，出现ValueError: Tensor Tensor("regression/concat:0", shape=(?, ?, 4), dtype=float32) is not an element of this graph.都是表明模型导入时图混乱，需要固定图

解决方式如下：

方法一：

init.py
加入detector.__graph = tf.get_default_graph()

from imageai.Detection import ObjectDetection
from flask import Flask
import tensorflow as tf
# 省略
detector = ObjectDetection()
detector.setModelTypeAsRetinaNet()
detector.setModelPath("./people/app/image_recognition/resnet50_coco_best_v2.0.1.h5")
detector.loadModel(detection_speed='fast')
detector.__graph = tf.get_default_graph()   # <<<<<<<<<<<<<<<保存模型图，防止模型混乱
# 省略
app = Flask(__name__)
app.config.from_object("food.config")

views.py
加入with detector.__graph.as_default():

### 省略
from .. import app, detector

def cutting_image(filename):
	with detector.__graph.as_default():   # <<<<<<<<<<<<<<<<<<<<<添加
	    custom_objects = detector.CustomObjects(bowl=True)
	    output_objects_array, detections, detected_objects_image_array = detector.detectCustomObjectsFromImage(input_image=filename, 
	        output_type = 'array',
	        minimum_percentage_probability=27,
	        extract_detected_objects = True,
	        custom_objects=custom_objects)

@xxx.route('/xxxx',methods=['POST'])
def get_msg():
	# 省略
	msg = cutting_image(filename)
	# 省略
	return result

方法二

views.py
detectCustomObjectsFromImage()自带参数thread_safe，设置为True则表示允许多线程下运行，
比较简便，直接用着方法就行了

def cutting_image(filename):
    custom_objects = detector.CustomObjects(bowl=True)
    output_objects_array, detections, detected_objects_image_array = detector.detectCustomObjectsFromImage(input_image=filename, 
        output_type = 'array',
        minimum_percentage_probability=27,
        extract_detected_objects = True,
        thread_safe=True,       # <<<<<<<<<<<<<<<<<<<<<<允许多线程
        custom_objects=custom_objects)

程序启动，输出如下所示，这里应该没什么问题，不用细看

现在开始调用接口

r = requests.post("http://127.0.0.1:5000/xxxx", data=body)

返回结果如下：

2019-12-26 11:29:01.091649: W ./tensorflow/core/grappler/optimizers/graph_optimizer_stage.h:241] Failed to run optimizer ArithmeticOptimizer, stage RemoveStackStridedSliceSameAxis node nms/strided_slice. Error: Pack node (boxes/stack) axis attribute is out of bounds: 2
2019-12-26 11:29:01.765712: W ./tensorflow/core/grappler/optimizers/graph_optimizer_stage.h:241] Failed to run optimizer ArithmeticOptimizer, stage RemoveStackStridedSliceSameAxis node nms/strided_slice. Error: Pack node (boxes/stack) axis attribute is out of bounds: 2
2019-12-26 11:29:02,142:ERROR:food.app:Exception on /recommend/recognition_food [POST]
Traceback (most recent call last):
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/imageai/Detection/__init__.py", line 717, in detectCustomObjectsFromImage
    _, _, detections = model.predict_on_batch(np.expand_dims(image, axis=0))
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/keras/engine/training.py", line 1580, in predict_on_batch
    outputs = self.predict_function(ins)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/tensorflow/python/keras/backend.py", line 3076, in __call__
    run_metadata=self.run_metadata)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 1439, in __call__
    run_metadata_ptr)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/errors_impl.py", line 528, in __exit__
    c_api.TF_GetCode(self.status.status))
tensorflow.python.framework.errors_impl.FailedPreconditionError: Error while reading resource variable bn2a_branch2b/gamma from Container: localhost. This could mean that the variable was uninitialized. Not found: Container localhost does not exist. (Could not find resource: localhost/bn2a_branch2b/gamma)
         [[{{node bn2a_branch2b/ReadVariableOp}}]]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 2446, in wsgi_app
    response = self.full_dispatch_request()
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1951, in full_dispatch_request
    rv = self.handle_user_exception(e)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1820, in handle_user_exception
    reraise(exc_type, exc_value, tb)
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/_compat.py", line 39, in reraise
    raise value
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1949, in full_dispatch_request
    rv = self.dispatch_request()
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/flask/app.py", line 1935, in dispatch_request
    return self.view_functions[rule.endpoint](**req.view_args)
  File "/home/yolanda/MyYolanda/user_analyze/food_recommend/project/food/app/image_recognition/views.py", line 198, in get_fod_msg
    all_foods_msg = recognition_all_food(path_open=image_path)
  File "/home/yolanda/MyYolanda/user_analyze/food_recommend/project/food/app/image_recognition/views.py", line 140, in recognition_all_food
    images_base64 = cutting_image(path_open)
  File "/home/yolanda/MyYolanda/user_analyze/food_recommend/project/food/app/image_recognition/views.py", line 45, in cutting_image
    custom_objects=custom_objects)           # 指定所要检测的物品类型，默认为全部
  File "/home/yolanda/anaconda3/lib/python3.7/site-packages/imageai/Detection/__init__.py", line 920, in detectCustomObjectsFromImage
    "Ensure you specified correct input image, input type, output type and/or output image path ")
ValueError: Ensure you specified correct input image, input type, output type and/or output image path

里面主要是这么个错误tensorflow.python.framework.errors_impl.FailedPreconditionError，提示是说读取错误

解决方式

因为我没有gpu，是用cpu的跑的，我怀疑有这方面原因

于是我到google colab上选择gpu进行测试，（google提供的免费云服务，免费提供gpu）

flask多线程下是能正常使用imageai的，输出结果正常

接着我还是在google colab上取消了gpu再次进行测试，发现结果依旧正常

对比了我本地和google colab里面依赖坏境，我keras版本较高

对此，我解决的方法是将keras降为2.2.5版本

pip install keras==2.2.5

问题解决

weixin_43412231

关注

1
点赞
踩
1

收藏

觉得还不错? 一键收藏
4
评论
Imageai用flask多线程部署出错

环境简要说明python3.7设计主要相关依赖包版本tensorflow = 1.15.0keras = 2.3.1flask = 1.1.1imageai = 2.1.5文件说明init.pyfrom imageai.Detection import ObjectDetectionfrom flask import Flaskimport tensorflow as ...
复制链接

扫一扫

专栏目录