本文主要參考下面兩篇博文,並在部分細節處作了修改。html
https://blog.csdn.net/linolzhang/article/details/97833354linux
1、數據集準備git
(訓練集驗證集測試集的數據分別準備)github
一、標註數據集json
大多數人會用labelme來標註數據集,而後用labelme將每張標註圖片都生成一個json文件。labelme教程網上不少,這裏再也不贅述。api
本人因爲原圖的標註目標很小,用labelme標註未免不精確,因此先用PS手動標註後再寫代碼把標註圖轉換成了labelme格式的json文件。服務器
結果如圖:app
二、將這些json文件轉換成coco格式ide
這一步我使用以下代碼可成功轉換。
# -*- coding:utf-8 -*- import os, sys import argparse import json import matplotlib.pyplot as plt import skimage.io as io from labelme import utils import numpy as np import glob import PIL.Image class MyEncoder(json.JSONEncoder): def default(self, obj): if isinstance(obj, np.integer): return int(obj) elif isinstance(obj, np.floating): return float(obj) elif isinstance(obj, np.ndarray): return obj.tolist() else: return super(MyEncoder, self).default(obj) class labelme2coco(object): def __init__(self, labelme_json=[], save_json_path='./tran.json'): ''' :param labelme_json: 全部labelme的json文件路徑組成的列表 :param save_json_path: json保存位置 ''' self.labelme_json = labelme_json self.save_json_path = save_json_path self.images = [] self.categories = [] self.annotations = [] # self.data_coco = {} self.label = [] self.annID = 1 self.height = 0 self.width = 0 self.save_json() def data_transfer(self): for num, json_file in enumerate(self.labelme_json): with open(json_file, 'r') as fp: data = json.load(fp) # 加載json文件 self.images.append(self.image(data, num)) for shapes in data['shapes']: label = shapes['label'] if label not in self.label: self.categories.append(self.categorie(label)) self.label.append(label) points = shapes['points'] # 這裏的point是用rectangle標註獲得的,只有兩個點,須要轉成四個點 points.append([points[0][0], points[1][1]]) points.append([points[1][0], points[0][1]]) self.annotations.append(self.annotation(points, label, num)) self.annID += 1 def image(self, data, num): image = {} #img = utils.img_b64_to_arr(data['imageData']) # 解析原圖片數據 # img=io.imread(data['imagePath']) # 經過圖片路徑打開圖片 # img = cv2.imread(data['imagePath'], 0) # height, width = img.shape[:2] height = data['imageHeight'] width = data['imageWidth'] image['height'] = height image['width'] = width image['id'] = num + 1 image['file_name'] = data['imagePath'].split('/')[-1] self.height = height self.width = width return image def categorie(self, label): categorie = {} categorie['supercategory'] = 'Cancer' categorie['id'] = len(self.label) + 1 # 0 默認爲背景 categorie['name'] = label return categorie def annotation(self, points, label, num): annotation = {} annotation['segmentation'] = [list(np.asarray(points).flatten())] annotation['iscrowd'] = 0 annotation['image_id'] = num + 1 # annotation['bbox'] = str(self.getbbox(points)) # 使用list保存json文件時報錯(不知道爲何) # list(map(int,a[1:-1].split(','))) a=annotation['bbox'] 使用該方式轉成list annotation['bbox'] = list(map(float, self.getbbox(points))) annotation['area'] = annotation['bbox'][2] * annotation['bbox'][3] # annotation['category_id'] = self.getcatid(label) annotation['category_id'] = self.getcatid(label) # 注意,源代碼默認爲1 annotation['id'] = self.annID return annotation def getcatid(self, label): for categorie in self.categories: if label == categorie['name']: return categorie['id'] return 1 def getbbox(self, points): # img = np.zeros([self.height,self.width],np.uint8) # cv2.polylines(img, [np.asarray(points)], True, 1, lineType=cv2.LINE_AA) # 畫邊界線 # cv2.fillPoly(img, [np.asarray(points)], 1) # 畫多邊形 內部像素值爲1 polygons = points mask = self.polygons_to_mask([self.height, self.width], polygons) return self.mask2box(mask) def mask2box(self, mask): '''從mask反算出其邊框 mask:[h,w] 0、1組成的圖片 1對應對象,只需計算1對應的行列號(左上角行列號,右下角行列號,就能夠算出其邊框) ''' # np.where(mask==1) index = np.argwhere(mask == 1) rows = index[:, 0] clos = index[:, 1] # 解析左上角行列號 left_top_r = np.min(rows) # y left_top_c = np.min(clos) # x # 解析右下角行列號 right_bottom_r = np.max(rows) right_bottom_c = np.max(clos) # return [(left_top_r,left_top_c),(right_bottom_r,right_bottom_c)] # return [(left_top_c, left_top_r), (right_bottom_c, right_bottom_r)] # return [left_top_c, left_top_r, right_bottom_c, right_bottom_r] # [x1,y1,x2,y2] return [left_top_c, left_top_r, right_bottom_c - left_top_c, right_bottom_r - left_top_r] # [x1,y1,w,h] 對應COCO的bbox格式 def polygons_to_mask(self, img_shape, polygons): mask = np.zeros(img_shape, dtype=np.uint8) mask = PIL.Image.fromarray(mask) xy = list(map(tuple, polygons)) PIL.ImageDraw.Draw(mask).polygon(xy=xy, outline=1, fill=1) mask = np.array(mask, dtype=bool) return mask def data2coco(self): data_coco = {} data_coco['images'] = self.images data_coco['categories'] = self.categories data_coco['annotations'] = self.annotations return data_coco def save_json(self): self.data_transfer() self.data_coco = self.data2coco() # 保存json文件 json.dump(self.data_coco, open(self.save_json_path, 'w'), indent=4, cls=MyEncoder) # indent=4 更加美觀顯示 if __name__ == '__main__': src_folder = os.path.abspath(sys.argv[1]) # load src - join json labelme_json = glob.glob(src_folder + '/*.json') labelme2coco(labelme_json, sys.argv[2])
在運行這個代碼時,只有把全部須要的模塊都安裝在anaconda當時安裝labelme的那個虛擬環境下才能運行成功。
2、環境搭建(linux)
一、建立pytorch環境
conda create --name maskrcnn_benchmark source activate maskrcnn_benchmark #全部模塊的安裝都在此虛擬環境下 conda install ipython pip install ninja yacs cython matplotlib pyqt5 conda install pytorch-nightly torchvision=0.2.1 cudatoolkit=9.0
上面的步驟執行完以後還要離線安裝torch1.0.1。由於某種牆的存在,在線下載torch不太容易實現,國內鏡像源又沒有1.0.1這個版本。而通過博主長期的踩坑發現torch1.0.1和torchvision=0.2.1加上numpy1.17纔是可用組合。這是torch1.0.1的下載連接: http://download.pytorch.org/whl/cu100/torch-1.0.1-cp36-cp36m-linux_x86_64.whl,建議直接迅雷下載。下載完成後,cd到模塊所在目錄而後pip install torch-1.0.1-cp36-cp36m-linux_x86_64.whl便可。(本人的python是3.6,請酌情修改下載連接)
二、安裝cocoapi及apex
export INSTALL_DIR=$PWD # install pycocotools git clone https://github.com/cocodataset/cocoapi.git cd cocoapi/PythonAPI python setup.py build_ext install # install apex cd $INSTALL_DIR git clone https://github.com/NVIDIA/apex.git cd apex python setup.py install --cuda_ext --cpp_ext
三、編譯模型代碼
# install PyTorch Detection cd $INSTALL_DIR #maskrcnn-benchmark #git clone https://github.com/facebookresearch/maskrcnn-benchmark.git git clone https://github.com/zjhuang22/maskscoring_rcnn cd maskscoring_rcnn python setup.py build develop
3、訓練前的準備
一、數據和預訓練模型準備
在下載的maskscoring_rcnn中新建一個datasets目錄,可按以下結構放置你的json文件和原始圖像
─ datasets └── annotations ├── coco_train.json └── coco_test.json └── coco_train #該文件夾放置訓練集的原始圖像 └── coco_test #該文件夾放置測試集的原始圖像
另外,maskscoring_rcnn的pretrained_models目錄下須要放置R-101.pkl和R-50.pkl這兩個預訓練模型,若是服務器連了網,在開始訓練模型以前會自動下載這兩個模型,若是服務器沒有網就須要手動下載放到pretrained_models下了。做者在GitHub也放了有這些模型的百度網盤連接。
二、修改參數
(1)修改 maskscoring_rcnn/configs
目錄下的配置文件,選擇其中的 e2e_ms_rcnn_R_50_FPN_1x.yaml
訓練腳本,修改以下:
MODEL: META_ARCHITECTURE: "GeneralizedRCNN" WEIGHT: "catalog://ImageNetPretrained/MSRA/R-50" PRETRAINED_MODELS: 'pretrained_models' DATASETS: TRAIN: ("coco_train_xxx",) # 1.設置訓練驗證集,名字能夠隨意起,和其餘配置文件對應便可。 TEST: ("coco_val_xxx",)
……(省略數行)
SOLVER:
BASE_LR: 0.002 #設置基礎學習率,原爲0.02
WEIGHT_DECAY: 0.0001
STEPS: (60000, 80000)
MAX_ITER: 5000 #2.設置最大迭代次數,可根據圖片數量酌情增減,改小也能夠更快看到結果。原爲90000
(2)修改 maskscoring_rcnn/maskrcnn_benchmark/config
下的 paths_catalog.py
文件:
DATASETS = { "coco_2014_train": ( "coco/train2014", "coco/annotations/instances_train2014.json",), "coco_2014_val": ("coco/val2014", "coco/annotations/instances_val2014.json"), "coco_2014_minival": ( "coco/val2014", "coco/annotations/instances_minival2014.json", ), "coco_2014_valminusminival": ( "coco/val2014", "coco/annotations/instances_valminusminival2014.json", ), #添加本身的數據集路徑信息,在相應的代碼段後面添加兩行便可 "coco_train_xxx": ("coco_mydata_train", "annotations/coco_mydata_train.json"), "coco_val_xxx": ("coco_mydata_test", "annotations/coco_mydata_test.json"), }
(3)修改 maskscoring_rcnn/maskrcnn_benchmark/config
下的 defaults.py
配置文件:
# Size of the smallest side of the image during training
_C.INPUT.MIN_SIZE_TRAIN = 800 # (800,)訓練集中圖片的最小邊長,酌情修改
# Maximum size of the side of the image during training
_C.INPUT.MAX_SIZE_TRAIN = 1333 #訓練集中圖片的最大邊長,酌情修改
# Size of the smallest side of the image during testing
_C.INPUT.MIN_SIZE_TEST = 800 #測試集中圖片的最小邊長,酌情修改
# Maximum size of the side of the image during testing
_C.INPUT.MAX_SIZE_TEST = 1333 #測試集中圖片的最大邊長,酌情修改
……省略數行……
_C.MODEL.ROI_BOX_HEAD.NUM_CLASSES = 3 # 修改分類數量,coco對應81(80+1),注意1加的是背景 _C.SOLVER.BASE_LR = 0.005 # 修改學習率,默認爲0.001 _C.SOLVER.CHECKPOINT_PERIOD = 1000 # 修改check point數量,根據須要自定義 _C.SOLVER.IMS_PER_BATCH = 1 # 修改batch size,默認16 _C.TEST.IMS_PER_BATCH = 1 # 修改test batch size,默認8 _C.OUTPUT_DIR = "weights/" # 設置模型保存路徑(對應自定義文件夾)
4、開始訓練
到maskscoring_rcnn
所在目錄下執行:
python tools/train_net.py --config-file configs/e2e_ms_rcnn_R_50_FPN_1x.yaml
python tools/test_net.py --config-file configs/e2e_ms_rcnn_R_50_FPN_1x.yaml
在models裏面能夠查看訓練日誌。
一、修改maskscoring_rcnn/configs
路徑下的對應的yaml
文件的權重路徑。
MODEL: META_ARCHITECTURE: "GeneralizedRCNN" WEIGHT: "weights/model_0005000.pth" # 訓練好的模型路徑 BACKBONE: CONV_BODY: "R-50-FPN" OUT_CHANNELS: 256
二、修改maskscoring_rcnn/demo
路徑下的 predictor.py
文件,添加類別信息。這個文件在原來的demo目錄下是沒有的,從mask rcnn benchmark的demo文件下複製過來便可。
class COCODemo(object): # COCO categories for pretty print CATEGORIES = [ "__background", "cla_a",#根據本身的數據集修改類別信息 "cla_b", "cla_c", ]
三、在maskscoring_rcnn/demo
下新建 predict.py,用於預測。
#!/usr/bin/env python # coding=UTF-8 import os, sys import numpy as np import cv2 from maskrcnn_benchmark.config import cfg from predictor import COCODemo # 1.修改後的配置文件 config_file = "configs/e2e_ms_rcnn_R_50_FPN_1x.yaml" # 2.配置 cfg.merge_from_file(config_file) # merge配置文件 cfg.merge_from_list(["MODEL.MASK_ON", True]) # 打開mask開關 cfg.merge_from_list(["MODEL.DEVICE", "cuda"]) # or設置爲CPU ["MODEL.DEVICE", "cpu"] #cfg.merge_from_list(["MODEL.DEVICE", "cpu"]) coco_demo = COCODemo( cfg, min_image_size=800, confidence_threshold=0.5, # 3.設置置信度 ) if __name__ == '__main__': in_folder = './datasets/test_images/' out_folder = './datasets/test_images_out/' if not os.path.exists(out_folder): os.makedirs(out_folder) for file_name in os.listdir(in_folder): if not file_name.endswith(('jpg', 'png')): continue # load file img_path = os.path.join(in_folder, file_name) image = cv2.imread(img_path) # method1. 直接獲得opencv圖片結果 #predictions = coco_demo.run_on_opencv_image(image) #save_path = os.path.join(out_folder, file_name) #cv2.imwrite(save_path, predictions) # method2. 獲取預測結果 predictions = coco_demo.compute_prediction(image) top_predictions = coco_demo.select_top_predictions(predictions) # draw img = coco_demo.overlay_boxes(image, top_predictions) img = coco_demo.overlay_mask(img, predictions) img = coco_demo.overlay_class_names(img, top_predictions) save_path = os.path.join(out_folder, file_name) cv2.imwrite(save_path, img) # print results boxes = top_predictions.bbox.numpy() labels = top_predictions.get_field("labels").numpy() #label = labelList[np.argmax(scores)] scores = top_predictions.get_field("scores").numpy() masks = top_predictions.get_field("mask").numpy() for i in range(len(boxes)): print('box:', i, ' label:', labels[i]) x1,y1,x2,y2 = [round(x) for x in boxes[i]] # = map(int, boxes[i]) print('x1,y1,x2,y2:', x1,y1,x2,y2)
四、運行程序。
python demo/predict.py
在運行的過程當中會報錯找不到文件或者沒法導入相關的庫,此時把相應的文件從 mask rcnn benchmark 對應的文件夾複製過來便可。具體操做可參考:http://www.javashuo.com/article/p-cmaoojcw-bk.html
成功截圖以下