import os
import sys
import random
import math
import re
import time
import numpy as np
import cv2
import matplotlib
import matplotlib.pyplot as plt
from PIL import Image # 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.config import Config
from mrcnn import utils
import mrcnn.model as modellib
from mrcnn import visualize
from mrcnn.model import log #%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) iter_num=0

  

Configurations

class ShapesConfig(Config):
"""Configuration for training on the toy shapes dataset.
Derives from the base Config class and overrides values specific
to the toy shapes dataset.
"""
# Give the configuration a recognizable name
NAME = "shapes" # Train on 1 GPU and 8 images per GPU. We can put multiple images on each
# GPU because the images are small. Batch size is 8 (GPUs * images/GPU).
GPU_COUNT = 2
IMAGES_PER_GPU = 1 #这里我用了两个GPU # Number of classes (including background)
NUM_CLASSES = 1 + 1 # background + 1 shapes # Use small images for faster training. Set the limits of the small side
# the large side, and that determines the image shape.
IMAGE_MIN_DIM = 1080
IMAGE_MAX_DIM = 1920 # Use smaller anchors because our image and objects are small
RPN_ANCHOR_SCALES = (8*6, 16*6, 32*6, 64*6, 128*6) # anchor side in pixels # Reduce training ROIs per image because the images are small and have
# few objects. Aim to allow ROI sampling to pick 33% positive ROIs.
TRAIN_ROIS_PER_IMAGE = 32 # Use a small epoch since the data is simple
STEPS_PER_EPOCH = 100 # use small validation steps since the epoch is small
VALIDATION_STEPS = 5 config = ShapesConfig()
config.display()

  Notebook Preference

def get_ax(rows=1, cols=1, size=8):
"""Return a Matplotlib Axes array to be used in
all visualizations in the notebook. Provide a
central point to control graph sizes. Change the default size attribute to control the size
of rendered images
"""
_, ax = plt.subplots(rows, cols, figsize=(size*cols, size*rows))
return ax

  Dataset

class DrugDataset(utils.Dataset):

    #得到该图中有多少个实例(物体)
def get_obj_index(self, image):
n = np.max(image)
return n
#解析labelme中得到的yaml文件,从而得到mask每一层对应的实例标签
def from_yaml_get_class(self,image_id):
info=self.image_info[image_id]
with open(info['yaml_path']) as f:
temp=yaml.load(f.read())
labels=temp['label_names']
del labels[0]
return labels
#重新写draw_mask
def draw_mask(self, num_obj, mask, image):
info = self.image_info[image_id]
for index in range(num_obj):
for i in range(info['width']):
for j in range(info['height']):
at_pixel = image.getpixel((i, j))
if at_pixel == index + 1:
mask[j, i, index] =1
return mask
#重新写load_shapes,里面包含自己的自己的类别(我的是box、column、package、fruit四类)
#并在self.image_info信息中添加了path、mask_path 、yaml_path
def load_shapes(self, count, height, width, img_floder, mask_floder, imglist,dataset_root_path):
"""Generate the requested number of synthetic images.
count: number of images to generate.
height, width: the size of the generated images.
"""
# Add classes
self.add_class("shapes", 1, "box") for i in range(count):
filestr = imglist[i].split(".")[0]
filestr = filestr.split("_")[0]
mask_path = mask_floder + "/" + filestr + ".png"
yaml_path=dataset_root_path+filestr+"rgb_"+"_json/info.yaml"
self.add_image("shapes", image_id=i, path=img_floder + "/"+imglist[i],
width=width, height=height, mask_path=mask_path,yaml_path=yaml_path)
#重写load_mask
def load_mask(self, image_id):
"""Generate instance masks for shapes of the given image ID.
"""
global iter_num
info = self.image_info[image_id]
count = 1 # number of object
img = Image.open(info['mask_path'])
num_obj = self.get_obj_index(img)
mask = np.zeros([info['height'], info['width'], num_obj], dtype=np.uint8)
mask = self.draw_mask(num_obj, mask, img)
occlusion = np.logical_not(mask[:, :, -1]).astype(np.uint8)
for i in range(count - 2, -1, -1):
mask[:, :, i] = mask[:, :, i] * occlusion
occlusion = np.logical_and(occlusion, np.logical_not(mask[:, :, i]))
labels=[]
labels=self.from_yaml_get_class(image_id)
labels_form=[]
for i in range(len(labels)):
if labels[i].find("box")!=-1:
#print "box"
labels_form.append("box")
#elif labels[i].find("column")!=-1:
#print "column"
# labels_form.append("column")
#elif labels[i].find("package")!=-1:
#print "package"
# labels_form.append("package")
#elif labels[i].find("fruit")!=-1:
#print "fruit"
# labels_form.append("fruit")
class_ids = np.array([self.class_names.index(s) for s in labels_form])
return mask, class_ids.astype(np.int32)

  基础设置

#基础设置
dataset_root_path="/mnt/disk2/zhouqiang/Mask_RCNN/data/train_01_01/"
img_floder = dataset_root_path+"rgb"
mask_floder = dataset_root_path+"mask"
#yaml_floder = dataset_root_path
imglist = os.listdir(img_floder)
count = len(imglist)
width = 1920
height = 1080 #train与val数据集准备
dataset_train = DrugDataset()
dataset_train.load_shapes(count, 1080, 1920, img_floder, mask_floder, imglist,dataset_root_path)
dataset_train.prepare() dataset_val = DrugDataset()
dataset_val.load_shapes(count, 1080, 1920, img_floder, mask_floder, imglist,dataset_root_path)
dataset_val.prepare()

  Create Model

# Create model in training mode
model = modellib.MaskRCNN(mode="training", config=config,
model_dir=MODEL_DIR)

 

# Which weights to start with?
init_with = "coco" # imagenet, coco, or last if init_with == "imagenet":
model.load_weights(model.get_imagenet_weights(), by_name=True)
elif init_with == "coco":
# Load weights trained on MS COCO, but skip layers that
# are different due to the different number of classes
# See README for instructions to download the COCO weights
model.load_weights(COCO_MODEL_PATH, by_name=True,
exclude=["mrcnn_class_logits", "mrcnn_bbox_fc",
"mrcnn_bbox", "mrcnn_mask"])
elif init_with == "last":
# Load the last model you trained and continue training
model.load_weights(model.find_last(), by_name=True)

  

# Fine tune all layers
# Passing layers="all" trains all layers. You can also
# pass a regular expression to select which layers to
# train by name pattern.
model.train(dataset_train, dataset_val,
learning_rate=config.LEARNING_RATE / 10,
epochs=50,
layers="all")

  

 

MaskRCNN 奔跑自己的数据的更多相关文章

  1. labelme2coco问题:TypeError: Object of type 'int64' is not JSON serializable

    最近在做MaskRCNN 在自己的数据(labelme)转为COCOjson格式遇到问题:TypeError: Object of type 'int64' is not JSON serializa ...

  2. Android动画之仿美团加载数据等待时,小人奔跑进度动画对话框(附顺丰快递员奔跑效果)

    Android动画之仿美团加载数据等待时,小人奔跑进度动画对话框(附顺丰快递员奔跑效果) 首句依然是那句老话,你懂得! finddreams :(http://blog.csdn.net/finddr ...

  3. 奔跑的歌颂 diskgenius 找回了20G数据

    2.0同学家的电脑不慎重装系统,结果默认重新分区.其他倒没什么数据,就是几千张记录孩子成长的照片最为珍贵.为了找回数据,用U盘启动,使用Diskgenius全部找回,在此奔歌一下.

  4. mask-rcnn代码解读(四):rpn_feature_maps数据的处理

    此处模拟 rpn_feature_maps数据的处理,最终得到rpn_class_logits, rpn_class, rpn_bbox. 代码如下: import numpy as np'''层与层 ...

  5. MVC5 网站开发之三 数据存储层功能实现

    数据存储层在项目Ninesky.DataLibrary中实现,整个项目只有一个类Repository.   目录 奔跑吧,代码小哥! MVC5网站开发之一 总体概述 MVC5 网站开发之二 创建项目 ...

  6. C#工业物联网和集成系统解决方案的技术路线(数据源、数据采集、数据上传与接收、ActiveMQ、Mongodb、WebApi、手机App)

    目       录 工业物联网和集成系统解决方案的技术路线... 1 前言... 1 第一章           系统架构... 3 1.1           硬件构架图... 3 1.2      ...

  7. jquery表格动态增删改及取数据绑定数据完整方案

    一 前言 上一篇Jquery遮罩插件,想罩哪就罩哪! 结尾的预告终于来了. 近期参与了一个针对内部员工个人信息收集的系统,其中有一个需求是在填写各个相关信息时,需要能动态的增加行当时公司有自己的解决方 ...

  8. 奔跑的xiaodao

    http://acm.hrbust.edu.cn/index.php?m=ProblemSet&a=showProblem&problem_id=2086 很明显的一个二分题目.因为要 ...

  9. 奔跑在Docker上的Spark

    转自:马踏飞燕--奔跑在Docker上的Spark 目录 为什么要在Docker上搭建Spark集群 网络拓扑 Docker安装及配置 ssh安装及配置 基础环境安装 Zookeeper安装及配置 H ...

随机推荐

  1. 转载:scala中的implicit

    掌握implicit的用法是阅读Spark源码的基础,也是学习Scala其它的开源框架的关键,implicit 可分为: 隐式参数 隐式转换类型 隐式调用函数 1.隐式参数 当我们在定义方法时,可以把 ...

  2. UDP 区别于 TCP 的特点

    TCP 我们了解得多了,所以今天我们站在 UDP 的角度,探讨一下 UDP 区别于 TCP 的特点. 1. 面向无连接 UDP 比 TCP 简单得多,不需要“三次握手”来建立连接,直接把内容发送出去. ...

  3. YII框架的行为

    一.什么是行为 行为,也称为 mixins,可以无须改变类继承关系即可增强一个已有的类的功能. 当一个对象或类被注入某些行为后,这个对象可以像访问自己定义的方法和属性一样访问注入进来的方法和属性. 二 ...

  4. docker容器连接

    前面我们实现了通过网络端口来访问运行在docker容器内的服务.下面我们来实现通过端口连接到一个docker容器 网络端口映射 我们创建了一个 python 应用的容器. runoob@runoob: ...

  5. [转] 修改sqlserver的数据库名、物理名称和逻辑文件名

    转载: https://blog.csdn.net/dym0080/article/details/81017777

  6. Django框架,python2和python3共存的情况下,创建Django项目

    一.python2和python3共存的情况下,直接使用 django-admin startproject Django-project 这个时候系统默认使用的是python2创建,可能由于pyth ...

  7. PluginWindowlessWin

    实际绘图发生在我的本机代码中的屏幕外目标上,每次刷新都会调用myplugin :: onWindowRefresh,它会将StretchBlt调用到插件的无窗口窗口,代码如下, FB::PluginW ...

  8. nodejs搜索包的流程

    执行npm install后,如果打包成功,会在当前目录下生成一个node_modules的文件夹,里面存放着我们所需的依赖包. 当需要引用时,例如: var math = require(" ...

  9. post请求头中常见content-type(非常重要)

    定义和用法 enctype 属性规定在发送到服务器之前应该如何对表单数据进行编码.默认地,表单数据会编码为 "application/x-www-form-urlencoded". ...

  10. Tomcat connection & session timeout settings

    # connection timeout for globle web application cat /home/soft/apache-tomcat-7.0.92/conf/server.xml ...