二：mask RCNN ——jupyter notebook mask rcnn

﹏ヽ暗。殇╰゛Y 2022-04-12 03:22 325阅读 0赞

默认你已经安装好环境了。以下教程完全遵照官网。

并且从官网下载好代码包了：[https://github.com/matterport/Mask\_RCNN][https_github.com_matterport_Mask_RCNN]，注意不是detrecton那个

直接到setup.py路径，也就是根路径

python3 setup.py install

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70][]

看下都干啥了：

WARNING:root:Fail load requirements file, so using default ones.
    running install
    running bdist_egg
    running egg_info
    creating mask_rcnn.egg-info
    writing mask_rcnn.egg-info/PKG-INFO
    writing dependency_links to mask_rcnn.egg-info/dependency_links.txt
    writing top-level names to mask_rcnn.egg-info/top_level.txt
    writing manifest file 'mask_rcnn.egg-info/SOURCES.txt'
    reading manifest file 'mask_rcnn.egg-info/SOURCES.txt'
    reading manifest template 'MANIFEST.in'
    writing manifest file 'mask_rcnn.egg-info/SOURCES.txt'
    installing library code to build/bdist.linux-x86_64/egg
    running install_lib
    running build_py
    creating build
    creating build/lib
    creating build/lib/mrcnn
    copying mrcnn/parallel_model.py -> build/lib/mrcnn
    copying mrcnn/model.py -> build/lib/mrcnn
    copying mrcnn/utils.py -> build/lib/mrcnn
    copying mrcnn/visualize.py -> build/lib/mrcnn
    copying mrcnn/config.py -> build/lib/mrcnn
    copying mrcnn/__init__.py -> build/lib/mrcnn
    creating build/bdist.linux-x86_64
    creating build/bdist.linux-x86_64/egg
    creating build/bdist.linux-x86_64/egg/mrcnn
    copying build/lib/mrcnn/parallel_model.py -> build/bdist.linux-x86_64/egg/mrcnn
    copying build/lib/mrcnn/model.py -> build/bdist.linux-x86_64/egg/mrcnn
    copying build/lib/mrcnn/utils.py -> build/bdist.linux-x86_64/egg/mrcnn
    copying build/lib/mrcnn/visualize.py -> build/bdist.linux-x86_64/egg/mrcnn
    copying build/lib/mrcnn/config.py -> build/bdist.linux-x86_64/egg/mrcnn
    copying build/lib/mrcnn/__init__.py -> build/bdist.linux-x86_64/egg/mrcnn
    byte-compiling build/bdist.linux-x86_64/egg/mrcnn/parallel_model.py to parallel_model.cpython-36.pyc
    byte-compiling build/bdist.linux-x86_64/egg/mrcnn/model.py to model.cpython-36.pyc
    byte-compiling build/bdist.linux-x86_64/egg/mrcnn/utils.py to utils.cpython-36.pyc
    byte-compiling build/bdist.linux-x86_64/egg/mrcnn/visualize.py to visualize.cpython-36.pyc
    byte-compiling build/bdist.linux-x86_64/egg/mrcnn/config.py to config.cpython-36.pyc
    byte-compiling build/bdist.linux-x86_64/egg/mrcnn/__init__.py to __init__.cpython-36.pyc
    creating build/bdist.linux-x86_64/egg/EGG-INFO
    copying mask_rcnn.egg-info/PKG-INFO -> build/bdist.linux-x86_64/egg/EGG-INFO
    copying mask_rcnn.egg-info/SOURCES.txt -> build/bdist.linux-x86_64/egg/EGG-INFO
    copying mask_rcnn.egg-info/dependency_links.txt -> build/bdist.linux-x86_64/egg/EGG-INFO
    copying mask_rcnn.egg-info/top_level.txt -> build/bdist.linux-x86_64/egg/EGG-INFO
    zip_safe flag not set; analyzing archive contents...
    creating dist
    creating 'dist/mask_rcnn-2.1-py3.6.egg' and adding 'build/bdist.linux-x86_64/egg' to it
    removing 'build/bdist.linux-x86_64/egg' (and everything under it)
    Processing mask_rcnn-2.1-py3.6.egg
    Copying mask_rcnn-2.1-py3.6.egg to /home/liutian/anaconda3/envs/mask1/lib/python3.6/site-packages
    Adding mask-rcnn 2.1 to easy-install.pth file
    
    Installed /home/liutian/anaconda3/envs/mask1/lib/python3.6/site-packages/mask_rcnn-2.1-py3.6.egg
    Processing dependencies for mask-rcnn==2.1
    Finished processing dependencies for mask-rcnn==2.1

**然后用jupyter lab打开demo.ipynb**

**很简单，就是在你的项目目录启动你建立的conda虚拟环境，如果你听不懂，那你估计就一个主环境，那你不用启动虚拟环境，进入项目目录。**

**输入：**

jupyter notebook

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 1][]

# 然后把地址复制到你的浏览器。 #

# 1 cell1 #

也就是第一个code cell。

import os
    import sys
    import random
    import math
    import numpy as np
    import skimage.io
    import matplotlib
    import matplotlib.pyplot as plt
    
    # Root directory of the project
    ROOT_DIR = os.path.abspath("../")
    
    # Import Mask RCNN
    sys.path.append(ROOT_DIR)  # To find local version of the library
    from mrcnn import utils
    import mrcnn.model as modellib
    from mrcnn import visualize
    # Import COCO config
    sys.path.append(os.path.join(ROOT_DIR, "samples/coco/"))  # To find local version
    import coco
    
    %matplotlib inline 
    
    # Directory to save logs and trained model
    MODEL_DIR = os.path.join(ROOT_DIR, "logs")
    
    # Local path to trained weights file
    COCO_MODEL_PATH = os.path.join(ROOT_DIR, "mask_rcnn_coco.h5")
    # Download COCO trained weights from Releases if needed
    if not os.path.exists(COCO_MODEL_PATH):
        utils.download_trained_weights(COCO_MODEL_PATH)
    
    # Directory of images to run detection on
    IMAGE_DIR = os.path.join(ROOT_DIR, "images")

用jupyter notebook打开：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 2][]

1 看一下./samples/coco/coco.py

2 下载coco工具包

3 注意下载作者改过的适用于python3的版本。[https://github.com/waleedka/coco][https_github.com_waleedka_coco]

4 然后改一下/pythonAPI下面的makefile

5 或者你不改成地址+python，你先激活环境再make，总之要保证你的实验安装等等都在该环境下。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 3][]

至于为什么作者没有直接改，因为这个coco工具包是python2，3通用的。所以作者propose a pull request

6 工具包也需要加入到路径，激活。方法（linux-python）就是用shell  进入PythonAPI 然后 命令行输入

make

make install

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 4][]

如果成功了，会在你的环境/lib里面出现pycocotools包

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 5][]

至于下载，作者的代码是可以自动下载的，但是默认下载2014.你也可以选择2017.

他会自动把coco权重下载到项目主目录，也就是/samples 所在目录。你也可以手动下载：

[https://github.com/matterport/Mask\_RCNN/releases][https_github.com_matterport_Mask_RCNN_releases]

coco图片：[http://cocodataset.org/\#download][http_cocodataset.org_download]如果要手动下载

然后看下coco.py代码，我选择用jupyter lab来肢解代码。先看下把lib加入到路径。这样你才能找到。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 6][]

然后看下coco.py怎么用，coco.py是所有的训练以及测试代码，建模不在这里。他负责处理你的命令：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 7][]

具体的处理代码在\_\_main\_\_

另外，权重的地址，以及val图片的地址如果不想放在根目录，可以改一下：

加入一行代码，更改图片地址：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 8][]

#  2 cell 2 #

第二个代码改了gpu的运算方式一个GPU运算一张图片

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 9][]

**这里改过了config，将会用于下一个建立模型，加载权重的代码段**

# 3 cell3 #

对于测试，我们需要把模型用原有的结构，但是新建立的参数重新建立一个，

权重就用之前的权重

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 10][]

下一个代码段是解决类别数目不连续的情况，这里掠过直接进如run object detection阶段

#  4. object detection #

用了新参数的config，建立的新的模型，对随机图片进行标注：

这里有两个未知，一个是verbose代表什么

另一个是r的内容

这需要看代码。

直接运行，我们可以得到：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 11][]

[https_github.com_matterport_Mask_RCNN]: https://github.com/matterport/Mask_RCNN
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70]: /images/20220412/50cee3e9e8fc474baecd358ebad0a15c.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 1]: /images/20220412/4a87fb8010f146aea941dd44f244b6f2.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 2]: /images/20220412/f2280fd09fd64a40b5285469c3aa66c9.png
[https_github.com_waleedka_coco]: https://github.com/waleedka/coco
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 3]: /images/20220412/7f977648d232475b9fcc55c51362ae64.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 4]: /images/20220412/609069b19ffe4515b0dbcd9215c541fb.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 5]: /images/20220412/2d8b5dd8c9a64b4fbb2844d0804f1be7.png
[https_github.com_matterport_Mask_RCNN_releases]: https://github.com/matterport/Mask_RCNN/releases
[http_cocodataset.org_download]: http://cocodataset.org/#download
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 6]: /images/20220412/e43adaa30c7d48a4a08a8d589ecd8415.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 7]: /images/20220412/84787554409b4911989ad922a6ce7fbf.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 8]: /images/20220412/4cc0a5178e74415b9c59a9363b75c710.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 9]: /images/20220412/aadadc43da794afc98e2ae0051ee7285.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 10]: /images/20220412/4fe89148748241cf86f2f67df9827cf0.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3UwMTMyNDk4NTM_size_16_color_FFFFFF_t_70 11]: /images/20220412/9b782cb3e11c48cabbc583f816d20877.png