|
|
这是机器未来的第4篇文章写在前面：
博客简介：专注AIoT领域，追逐未来时代的脉搏，记录路途中的技术成长！
专栏简介：记录博主从0到1掌握物体检测工作流的过程，具备自定义物体检测器的能力
面向人群：具备深度学习理论基础的学生或初级开发者
专栏计划：接下来会逐步发布跨入人工智能的系列博文，敬请期待
Python零基础快速入门系列
快速入门Python数据科学系列
人工智能开发环境搭建系列
机器学习系列
物体检测快速入门系列
自动驾驶物体检测系列
......
1. 概述
tensorflow object detection api一个框架，它可以很容易地构建、训练和部署对象检测模型，并且是一个提供了众多基于COCO数据集、Kitti数据集、Open Images数据集、AVA v2.1数据集和iNaturalist物种检测数据集上提供预先训练的对象检测模型集合。
kites_detections_output
tensorflow object detection api是目前最主流的目标检测框架之一，主流的目标检测模型如图所示：
snipaste20220513_0948282. 预置条件
为了顺利按照本手册安装tensroflow object detection api，请参考Windows部署Docker GPU深度学习开发环境安装必备的工具。
若自行创建安装条件，请确保已经满足以下条件
支持python3.8以上版本
支持cuda、cudnn（可选）
支持git
本手册使用docker运行环境。
3. 安装步骤
3.1 Docker环境
3.1.1 启动docker
启动docker桌面客户端，如图所示：
13.1.2 启动容器
在windows平台可以启动命令行工具或者windows terminal工具(App Store下载)，这里使用terminal工具。
输入以下命令，查看当前存在的images列表
PSC:\Users\xxxxx>dockerimages
REPOSITORYTAGIMAGEIDCREATEDSIZE
docker/getting-startedlatestbd9a9f7338985weeksago28.8MBtensorflow/tensorflow2.8.0-gpu-jupytercc9a9ae2a5af6weeksago5.99GB
可以看到之前安装的tensorflow-2.8.0-gpu-jupyter镜像，现在基于这个镜像启动容器
dockerrun--gpusall-itd-ve:/dockerdir/docker_work/:/home/zhou/-p8888:8888-p6006:6006--ipc=hostcc9a9ae2a5afjupyternotebook--no-browser--ip=0.0.0.0--allow-root--NotebookApp.token=--notebook-dir='/home/zhou/'命令释义：docker run：表示基于镜像启动容器 --gpus all：不加此选项,nvidia-smi命令会不可用 -i: 交互式操作。-t: 终端。-d:后台运行，需要使用【docker exec -it 容器id /bin/bash】进入容器 -v e:/dockerdir/docker_work/:/home/zhou/:将windows平台的e:/dockerdir/docker_work/目录映射到docker的ubuntu系统的/home/zhou/目录下，实现windows平台和docker系统的文件共享 -p 8888:8888 -p 6006:6006：表示将windows系统的8888、6006端口映射到docker的8888、6006端口，这两个端口分别为jupyter notebook和tensorboard的访问端口 --ipc=host:用于多个容器之间的通讯 cc9a9ae2a5af:tensorflow-2.8.0-gpu-jupyter镜像的IMAGE ID jupyter notebook --no-browser --ip=0.0.0.0 --allow-root --NotebookApp.token= --notebook-dir='/home/zhou/': docker开机启动命令，这里启动jupyter
启动vscode后，选择docker工具栏，在启动的容器上，右键选择附着到VsCode
23.1.4 更换docker容器ubuntu系统的安装源为国内源
在vscode软件界面上，选择【文件】-【打开文件夹】，选择根目录【/】，找到【/etc/apt/sources.list】,将ubuntu的安装源全部切换为aliyun源，具体操作为：将【archive.ubuntu.com】修改为【mirrors.aliyun.com】即可，修改后如下：
#Seehttp://help.ubuntu.com/community/UpgradeNotesforhowtoupgradeto
#newerversionsofthedistribution.
debhttp://mirrors.aliyun.com/ubuntu/focalmainrestricted
#deb-srchttp://mirrors.aliyun.com/ubuntu/focalmainrestricted
##Majorbugfixupdatesproducedafterthefinalreleasofthe
##distribution.
debhttp://mirrors.aliyun.com/ubuntu/focal-updatesmainrestricted
#deb-srchttp://mirrors.aliyun.com/ubuntu/focal-updatesmainrestricted
##N.B.softwarefromthisrepositoryisENTIRELYUNSUPPORTEDbytheUbuntu
##team.Also,pleasenotethatsoftwareinuniverseWILLNOTreceiveany
##revieworupdatesfromtheUbuntusecurityteam.
debhttp://mirrors.aliyun.com/ubuntu/focaluniverse
#deb-srchttp://mirrors.aliyun.com/ubuntu/focaluniverse
debhttp://mirrors.aliyun.com/ubuntu/focal-updatesuniverse
#deb-srchttp://mirrors.aliyun.com/ubuntu/focal-updatesuniverse
##N.B.softwarefromthisrepositoryisENTIRELYUNSUPPORTEDbytheUbuntu
##team,andmaynotbeunderafreelicence.Pleasesatisfyyourselfasto
##yourrightstousethesoftware.Also,pleasenotethatsoftwarein
##multiverseWILLNOTreceiveanyrevieworupdatesfromtheUbuntu
##securityteam.
debhttp://mirrors.aliyun.com/ubuntu/focalmultiverse
#deb-srchttp://mirrors.aliyun.com/ubuntu/focalmultiverse
debhttp://mirrors.aliyun.com/ubuntu/focal-updatesmultiverse
#deb-srchttp://mirrors.aliyun.com/ubuntu/focal-updatesmultiverse
##N.B.softwarefromthisrepositorymaynothavebeentestedas
##extensivelyasthatcontainedinthemainrelease,althoughitincludes
##newerversionsofsomeapplicationswhichmayprovideusefulfeatures.
##Also,pleasenotethatsoftwareinbackportsWILLNOTreceiveanyreview
##orupdatesfromtheUbuntusecurityteam.
debhttp://mirrors.aliyun.com/ubuntu/focal-backportsmainrestricteduniversemultiverse
#deb-srchttp://mirrors.aliyun.com/ubuntu/focal-backportsmainrestricteduniversemultiverse
##UncommentthefollowingtwolinestoaddsoftwarefromCanonical's
##'partner'repository.
##ThissoftwareisnotpartofUbuntu,butisofferedbyCanonicalandthe
##respectivevendorsasaservicetoUbuntuusers.
#debhttp://archive.canonical.com/ubuntufocalpartner
#deb-srchttp://archive.canonical.com/ubuntufocalpartner
debhttp://security.ubuntu.com/ubuntu/focal-securitymainrestricted
#deb-srchttp://security.ubuntu.com/ubuntu/focal-securitymainrestricted
debhttp://security.ubuntu.com/ubuntu/focal-securityuniverse
#deb-srchttp://security.ubuntu.com/ubuntu/focal-securityuniverse
debhttp://security.ubuntu.com/ubuntu/focal-securitymultiverse#deb-srchttp://security.ubuntu.com/ubuntu/focal-securitymultiverse
apt-getupdate;apt-get-finstall;apt-getupgrade
3.1.5 验证GPU是否加载成功（在电脑有Nvidia显卡的情况下）
输入nvidia-smi查看GPU使用情况，nvcc -V查询cuda版本
root@cc58e655b170:/home/zhou#nvidia-smi
TueMar2215:08:572022
+-----------------------------------------------------------------------------+
|NVIDIA-SMI470.85DriverVersion:472.47CUDAVersion:11.4|
|-------------------------------+----------------------+----------------------+
|GPUNamePersistence-M|Bus-IdDisp.A|VolatileUncorr.ECC|
|FanTempPerfPwr:Usage/Cap|Memory-Usage|GPU-UtilComputeM.|
|||MIGM.|
|===============================+======================+======================|
|0NVIDIAGeForce...Off|00000000:01:00.0Off|N/A|
|N/A48CP89W/N/A|153MiB/6144MiB|ERR!Default|
|||N/A|
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
|Processes:|
|GPUGICIPIDTypeProcessnameGPUMemory|
|IDIDUsage|
|=============================================================================|
|Norunningprocessesfound|
+-----------------------------------------------------------------------------+
root@cc58e655b170:/home/zhou#nvcc-V
nvcc:NVIDIA(R)Cudacompilerdriver
Copyright(c)2005-2021NVIDIACorporation
BuiltonSun_Feb_14_21:12:58_PST_2021
Cudacompilationtools,release11.2,V11.2.152Buildcuda_11.2.r11.2/compiler.29618528_0
从nvcc -V的日志，可以看出cuda版本为11.2
python-c"importtensorflowastf;print(tf.reduce_sum(tf.random.normal([1000,1000])))"
输出结果如下：
root@cc58e655b170:/usr#python-c"importtensorflowastf;print(tf.reduce_sum(tf.random.normal([1000,1000])))"
2022-03-2215:26:13.281719:Itensorflow/core/common_runtime/gpu/gpu_device.cc:1525]Createddevice/job:localhost/replica:0/task:0/device:GPU:0with3951MBmemory:->device:0,name:NVIDIAGeForceGTX1660Ti,pcibusid:0000:01:00.0,computecapability:7.5tf.Tensor(-2613.715,shape=(),dtype=float32)
从输出日志，可以看到GPU：NVIDIA GeForce GTX 1660 Ti已经加载到docker，cuDNN版本为7.5
3.2 Windows开发环境
同Docker环境，验证cuda和cuDNN安装情况。
3.3 下载tensorflow object detection api项目源码
在home/zhou目录下创建tensorflow的目录
cd/home/zhou;mkdirtensorflow;cdtensorflow
gitclonehttps://github.com/tensorflow/models.git
下载完毕后，默认文件名名称为models-master, 将文件名重命名为models,保持文件名和平台一致
mvmodels-matsermodels
如果网速不好，直接下载zip压缩包吧
3
下载完毕后的文档结构如图所示：
tensorflow/
└─models/
├─community/
├─official/
├─orbit/
├─research/└──...3.4 安装配置protobuf
Tensorflow对象检测API使用Protobufs来配置模型和训练参数。在使用框架之前，必须下载并编译Protobuf库。
cd/home/zhou
下载protobuf 这里下载的已经预编译好的protobuf
wget-chttps://github.com/protocolbuffers/protobuf/releases/download/v3.19.4/protoc-3.19.4-linux-x86_64.zip
解压 先执行mkdir protoc-3.19.4创建目录，然后执行unzip protoc-3.19.4-linux-x86_64.zip -d protoc-3.19.4/解压到制定目录protoc-3.19.4
root@cc58e655b170:/home/zhou#mkdirprotoc-3.19.4
root@cc58e655b170:/home/zhou#unzipprotoc-3.19.4-linux-x86_64.zip-dprotoc-3.19.4/
Archive:protoc-3.19.4-linux-x86_64.zip
creating:protoc-3.19.4/include/
creating:protoc-3.19.4/include/google/
creating:protoc-3.19.4/include/google/protobuf/
inflating:protoc-3.19.4/include/google/protobuf/wrappers.proto
inflating:protoc-3.19.4/include/google/protobuf/source_context.proto
inflating:protoc-3.19.4/include/google/protobuf/struct.proto
inflating:protoc-3.19.4/include/google/protobuf/any.proto
inflating:protoc-3.19.4/include/google/protobuf/api.proto
inflating:protoc-3.19.4/include/google/protobuf/descriptor.proto
creating:protoc-3.19.4/include/google/protobuf/compiler/
inflating:protoc-3.19.4/include/google/protobuf/compiler/plugin.proto
inflating:rotoc-3.19.4/include/google/protobuf/timestamp.proto
inflating:protoc-3.19.4/include/google/protobuf/field_mask.proto
inflating:protoc-3.19.4/include/google/protobuf/empty.proto
inflating:protoc-3.19.4/include/google/protobuf/duration.proto
inflating:protoc-3.19.4/include/google/protobuf/type.proto
creating:protoc-3.19.4/bin/
inflating:protoc-3.19.4/bin/protocinflating:protoc-3.19.4/readme.txt
配置protoc 在~/.bashrc文件的末尾添加如下代码
exportPATH=$PATH:/home/zhou/protoc-3.19.4/bin
执行如下命令，使其生效
source~/.bashrc
执行echo $PATH查看是否生效
root@cc58e655b170:/home/zhou/protoc-3.19.4/bin#echo$PATH/home/zhou/protoc-3.19.4/bin:/home/zhou/protoc-3.19.4/bin:/home/zhou/protoc-3.19.4/bin:/root/.vscode-server/bin/c722ca6c7eed3d7987c0d5c3df5c45f6b15e77d1/bin/remote-cli:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/home/zhou/protoc-3.19.4/bin
可以看到protoc的安装目录/home/zhou/protoc-3.19.4/bin已经添加到PATH了。
3.5 将proto后缀文件转换为python可识别格式
cd/home/zhou/tensorflow/models/research/
lsobject_detection/protos/
4
转换proto文件格式为python可识别序列化文件
protocobject_detection/protos/*.proto--python_out=.
lsobject_detection/protos/
53.6 安装coco api
从TensorFlow 2.x开始, pycocotools包被列为对象检测API的依赖项。理想情况下，这个包应该在安装对象检测API时安装，如下面安装对象检测API一节所述，但是由于各种原因，安装可能会失败，因此更简单的方法是提前安装这个包，在这种情况下，稍后的安装将被跳过。
pipinstallcythonpipinstallgit+https://github.com/philferriere/cocoapi.git#subdirectory=PythonAPI
默认指标是基于Pascal VOC评估中使用的那些指标。要使用COCO对象检测指标，在配置文件的eval_config消息中添加metrics_set: "coco_detection_metrics"。要使用COCO实例分割度量，在配置文件的eval_config消息中添加metrics_set: "coco_mask_metrics"。
3.7 安装object detection api
root@cc58e655b170:/home/zhou/tensorflow/models/research#pwd/home/zhou/tensorflow/models/research
cpobject_detection/packages/tf2/setup.py.python-mpipinstall--use-feature=2020-resolver.
安装过程会持续一段时间，安装完毕后，可以执行如下代码，测试安装是否完成。
pythonobject_detection/builders/model_builder_tf2_test.py
输出如下：
......
I032216:48:09.677789140205126002496efficientnet_model.py:144]round_filterinput=192output=384
I032216:48:10.876914140205126002496efficientnet_model.py:144]round_filterinput=192output=384
I032216:48:10.877072140205126002496efficientnet_model.py:144]round_filterinput=320output=640
I032216:48:11.294571140205126002496efficientnet_model.py:144]round_filterinput=1280output=2560
I032216:48:11.337533140205126002496efficientnet_model.py:454]BuildingmodelefficientnetwithparamsModelConfig(width_coefficient=2.0,depth_coefficient=3.1,resolution=600,dropout_rate=0.5,blocks=(BlockConfig(input_filters=32,output_filters=16,kernel_size=3,num_repeat=1,expand_ratio=1,strides=(1,1),se_ratio=0.25,id_skip=True,fused_conv=False,conv_type='depthwise'),BlockConfig(input_filters=16,output_filters=24,kernel_size=3,num_repeat=2,expand_ratio=6,strides=(2,2),se_ratio=0.25,id_skip=True,fused_conv=False,conv_type='depthwise'),BlockConfig(input_filters=24,output_filters=40,kernel_size=5,num_repeat=2,expand_ratio=6,strides=(2,2),se_ratio=0.25,id_skip=True,fused_conv=False,conv_type='depthwise'),BlockConfig(input_filters=40,output_filters=80,kernel_size=3,num_repeat=3,expand_ratio=6,strides=(2,2),se_ratio=0.25,id_skip=True,fused_conv=False,conv_type='depthwise'),BlockConfig(input_filters=80,output_filters=112,kernel_size=5,num_repeat=3,expand_ratio=6,strides=(1,1),se_ratio=0.25,id_skip=True,fused_conv=False,conv_type='depthwise'),BlockConfig(input_filters=112,output_filters=192,kernel_size=5,num_repeat=4,expand_ratio=6,strides=(2,2),se_ratio=0.25,id_skip=True,fused_conv=False,conv_type='depthwise'),BlockConfig(input_filters=192,output_filters=320,kernel_size=3,num_repeat=1,expand_ratio=6,strides=(1,1),se_ratio=0.25,id_skip=True,fused_conv=False,conv_type='depthwise')),stem_base_filters=32,top_base_filters=1280,activation='simple_swish',batch_norm='default',bn_momentum=0.99,bn_epsilon=0.001,weight_decay=5e-06,drop_connect_rate=0.2,depth_divisor=8,min_depth=None,use_se=True,input_channels=3,num_classes=1000,model_name='efficientnet',rescale_input=False,data_format='channels_last',dtype='float32')
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_create_ssd_models_from_config):33.12s
I032216:48:11.521103140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_create_ssd_models_from_config):33.12s
[OK]ModelBuilderTF2Test.test_create_ssd_models_from_config
[RUN]ModelBuilderTF2Test.test_invalid_faster_rcnn_batchnorm_update
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_invalid_faster_rcnn_batchnorm_update):0.0s
I032216:48:11.532667140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_invalid_faster_rcnn_batchnorm_update):0.0s
[OK]ModelBuilderTF2Test.test_invalid_faster_rcnn_batchnorm_update
[RUN]ModelBuilderTF2Test.test_invalid_first_stage_nms_iou_threshold
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_invalid_first_stage_nms_iou_threshold):0.0s
I032216:48:11.535152140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_invalid_first_stage_nms_iou_threshold):0.0s
[OK]ModelBuilderTF2Test.test_invalid_first_stage_nms_iou_threshold
[RUN]ModelBuilderTF2Test.test_invalid_model_config_proto
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_invalid_model_config_proto):0.0s
I032216:48:11.535965140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_invalid_model_config_proto):0.0s
[OK]ModelBuilderTF2Test.test_invalid_model_config_proto
[RUN]ModelBuilderTF2Test.test_invalid_second_stage_batch_size
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_invalid_second_stage_batch_size):0.0s
I032216:48:11.539124140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_invalid_second_stage_batch_size):0.0s
[OK]ModelBuilderTF2Test.test_invalid_second_stage_batch_size
[RUN]ModelBuilderTF2Test.test_session
[SKIPPED]ModelBuilderTF2Test.test_session
[RUN]ModelBuilderTF2Test.test_unknown_faster_rcnn_feature_extractor
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_unknown_faster_rcnn_feature_extractor):0.0s
I032216:48:11.542018140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_unknown_faster_rcnn_feature_extractor):0.0s
[OK]ModelBuilderTFTest.test_unknown_faster_rcnn_feature_extractor
[RUN]ModelBuilderTF2Test.test_unknown_meta_architecture
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_unknown_meta_architecture):0.0s
I032216:48:11.543226140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_unknown_meta_architecture):0.0s
[OK]ModelBuilderTF2Test.test_unknown_meta_architecture
[RUN]ModelBuilderTF2Test.test_unknown_ssd_feature_extractor
INFO:tensorflow:time(__main__.ModelBuilderTF2Test.test_unknown_ssd_feature_extractor):0.0s
I032216:48:11.545147140205126002496test_util.py:2373]time(__main__.ModelBuilderTF2Test.test_unknown_ssd_feature_extractor):0.0s
[OK]ModelBuilderTF2Test.test_unknown_ssd_feature_extractor
----------------------------------------------------------------------
Ran24testsin42.982s
OK(skipped=1)
看到结果为OK,则表示安装成功，接下来就可以开始物体检测之旅了。
《物体检测快速入门系列》快速导航：
物体检测快速入门系列(1)-基于Tensorflow2.x Object Detection API构建自定义物体检测器
物体检测快速入门系列(2)-Windows部署GPU深度学习开发环境
物体检测快速入门系列(3)-Windows部署Docker GPU深度学习开发环境
物体检测快速入门系列(4)-TensorFlow 2.x Object Detection API快速安装手册
|
本帖子中包含更多资源
您需要 登录 才可以下载或查看，没有账号？立即注册
×
|
窥视卡
雷达卡
发表于 2022-5-26 11:56
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
千斤顶
显身卡