数据读取在faster-rcnn源码里是比较简单的部分,但也是非常重要的部分,不了解数据,就不可能了解算法。

另一方面,由于python环境碎片话化,源码调用的库在你的电脑上如果碰巧(其实概率还蛮大,特别是windows下)不能用,完全可以用另外一种等价的方式取代。

一、图片读取就是把图片转化成矩阵,等待下一个流程进一步处理。图片读取要注意不是所有都是RGB顺序读取

1.cv2(OpenCV-Python :BGR顺序

img = cv2.imread(image_path)
h, w, c = image.shape
img = cv2.resize(img, (self.img_width, self.img_height))
img = img[:,:,::-1]#bgr转rgb

2.PIL(Pillow):

from PIL import Image
img = Image.open(img_path).convert('RGB')
img = Image.fromarray(img)

3.skimage(scikit-image):

from skimage import io
img=io.imread(image_path)from skimage import transform
img = transform.resize(img, (C, H * scale, W * scale), mode='reflect',anti_aliasing=False)

4.scipy.misc

import scipy.misc
#或scipy.ndimage.imread
img = scipy.misc.imread(image_path)

5.matplotlib

import matplotlib.image as mpimg
img = mpimg.imread(image_path)

前两种方式用得很多,后几种方式用的很少。

这些库很多,但万剑归功,读取基本都是调用PIL或libjpeg,详情看

Python的各种imread函数在实现方式和读取速度上有何区别?​www.zhihu.com

大致结论是:读取PIL最快,resize的话CV2最快。

二、标签读取

1.import xml.etree.ElementTree as ET

这个读取xml是标配

三、pascal_voc数据集

1.在/VOCdevkit/VOC2007/下有五个文件夹,faster-rcnn会用到其中三个:

(1)ImageSets:划分(split)数据,txt文件存文件名(不带后缀),划分得比较多,也比较细,有不同用途,faster-rcnn只要其中的 'ImageSets/Main/{0}.txt'.format(split)即可,这里split in ['train', 'test', 'val']或其他别的也可以。

每个txt文件每行一个文件名,一般一个文件几千行,对应几千个文件。

(2)JPEGImages:存放图片,文件名+.jpg即可。

(3)Annotations:存放标签,文件名+.xml即可。

xml格式为:

<annotation><folder>VOC2007</folder><filename>000001.jpg</filename><source><database>The VOC2007 Database</database><annotation>PASCAL VOC2007</annotation><image>flickr</image><flickrid>341012865</flickrid></source><owner><flickrid>Fried Camels</flickrid><name>Jinky the Fruit Bat</name></owner><size><width>353</width><height>500</height><depth>3</depth></size><segmented>0</segmented><object><name>dog</name><pose>Left</pose><truncated>1</truncated><difficult>0</difficult><bndbox><xmin>48</xmin><ymin>240</ymin><xmax>195</xmax><ymax>371</ymax></bndbox></object><object><name>person</name><pose>Left</pose><truncated>1</truncated><difficult>0</difficult><bndbox><xmin>8</xmin><ymin>12</ymin><xmax>352</xmax><ymax>498</ymax></bndbox></object>
</annotation>

只要关注object部分:size、name和bndbox即可,其他部分不用太关心。

四、pytorch的dataset

这篇文章讲得很清楚:

Elijha:Pytorch数据读取(Dataset, DataLoader, DataLoaderIter)​zhuanlan.zhihu.com

概括起来:

torch.utils.data.Dataset:

最基本的读取,负责表示数据集,两个方法__getitem__()和__len__()

torch.utils.data.DataLoader

在Dataset之上+batch+多线程+数据处理

torch.utils.data.dataloader.DataLoaderIter

DataLoader之上+迭代

五、部分源码

来自:facebookresearch/maskrcnn-benchmark

继承自最基础的torch.utils.data.Dataset
Dataset可以交给DataLoader,封装batchs和多线程
class PascalVOCDataset(torch.utils.data.Dataset):第一个是背景,其他是目标类别CLASSES = ("__background__ ","aeroplane","bicycle","bird","boat","bottle","bus","car","cat","chair","cow","diningtable","dog","horse","motorbike","person","pottedplant","sheep","sofa","train","tvmonitor",)def __init__(self, data_dir, split, use_difficult=False, transforms=None):
数据根目录self.root = data_dir
指定用哪个数据,test、val、train?self.image_set = split
要不要难以识别的图片?self.keep_difficult = use_difficult
图片处理self.transforms = transforms
对应三个文件加:self._annopath = os.path.join(self.root, "Annotations", "%s.xml")self._imgpath = os.path.join(self.root, "JPEGImages", "%s.jpg")
直接用Main文件夹self._imgsetpath = os.path.join(self.root, "ImageSets", "Main", "%s.txt")
初始化的时候就把文件名全读出来with open(self._imgsetpath % self.image_set) as f:self.ids = f.readlines()self.ids = [x.strip("n") for x in self.ids]
构造id到图片的字典,self.id_to_img_map = {k: v for k, v in enumerate(self.ids)}cls = PascalVOCDataset.CLASSES
类别到序号和序号到类别的两个字典self.class_to_ind = dict(zip(cls, range(len(cls))))self.categories = dict(zip(range(len(cls)), cls))
__getitem__是必须的def __getitem__(self, index):img_id = self.ids[index]
根据id号读取图片,用的是PIL来读img = Image.open(self._imgpath % img_id).convert("RGB")
获取目标框,裁剪target = self.get_groundtruth(index)target = target.clip_to_image(remove_empty=True)
做归一化,缩放等处理if self.transforms is not None:img, target = self.transforms(img, target)
target 前四个表示坐标,第五个值表示类别 return img, target, index
__len__也是必须的def __len__(self):return len(self.ids)def get_groundtruth(self, index):img_id = self.ids[index]
定位到根节点anno = ET.parse(self._annopath % img_id).getroot()
读取图片大小、目标类别、目标位置信息anno = self._preprocess_annotation(anno)
取出高宽信息height, width = anno["im_info"]
按xyxy顺序的坐标信息,这个地方要注意,并不是每个实现都是相同顺序,还是看读取的时候怎么读的,有的是yxyx顺序
BoxList是自定义的处理这几个不目标框的类target = BoxList(anno["boxes"], (width, height), mode="xyxy")
BoxList添加标签,难度target.add_field("labels", anno["labels"])target.add_field("difficult", anno["difficult"])return targetdef _preprocess_annotation(self, target):boxes = []gt_classes = []difficult_boxes = []TO_REMOVE = 1
只遍历object节点for obj in target.iter("object"):
遇到难以识别的,跳过也行,不是强烈要求,就跳过吧difficult = int(obj.find("difficult").text) == 1if not self.keep_difficult and difficult:continue
标签名称name = obj.find("name").text.lower().strip()
矩形坐标bb = obj.find("bndbox")# Make pixel indexes 0-based# Refer to "https://github.com/rbgirshick/py-faster-rcnn/blob/master/lib/datasets/pascal_voc.py#L208-L211"box = [bb.find("xmin").text,bb.find("ymin").text,bb.find("xmax").text,bb.find("ymax").text,]
两次map:
第一次是把文本转化成int,这里int是函数;
第二次是把int减去1,0-based
得到一个tuplebndbox = tuple(map(lambda x: x - TO_REMOVE, list(map(int, box))))
把这tuple添加进boxesboxes.append(bndbox)
把标签转化成类别idgt_classes.append(self.class_to_ind[name])difficult_boxes.append(difficult)循环结束,一张图片目标有一个或多个,但大小只有一个
看图片大小size = target.find("size")
把高宽从文本转化成int,得到一个tupleim_info = tuple(map(int, (size.find("height").text, size.find("width").text)))
把这张图片的信息用一个字典返回
并把坐标转化成floatres = {"boxes": torch.tensor(boxes, dtype=torch.float32),"labels": torch.tensor(gt_classes),"difficult": torch.tensor(difficult_boxes),"im_info": im_info,}return res

六、coco数据集

coco跟voc差不太多,但是标签不同,voc是单独的xml,而coco是一个json,种类更多,层次更多,数据信息也更多,大致像这样:

{"info": {"description": "COCO 2014 Dataset","url": "http://cocodataset.org","version": "1.0","year": 2014,"contributor": "COCO Consortium","date_created": "2017/09/01"},"licenses": [{"url": "http://creativecommons.org/licenses/by-nc-sa/2.0/","id": 1,"name": "Attribution-NonCommercial-ShareAlike License"}, {"url": "http://creativecommons.org/licenses/by-nc/2.0/","id": 2,"name": "Attribution-NonCommercial License"}, {"url": "http://creativecommons.org/licenses/by-nc-nd/2.0/","id": 3,"name": "Attribution-NonCommercial-NoDerivs License"}, {"url": "http://creativecommons.org/licenses/by/2.0/","id": 4,"name": "Attribution License"}, {"url": "http://creativecommons.org/licenses/by-sa/2.0/","id": 5,"name": "Attribution-ShareAlike License"}, {"url": "http://creativecommons.org/licenses/by-nd/2.0/","id": 6,"name": "Attribution-NoDerivs License"}, {"url": "http://flickr.com/commons/usage/","id": 7,"name": "No known copyright restrictions"}, {"url": "http://www.usa.gov/copyright.shtml","id": 8,"name": "United States Government Work"}],"categories": [{"supercategory": "person","id": 1,"name": "person"}, {"supercategory": "vehicle","id": 2,"name": "bicycle"}, {"supercategory": "vehicle","id": 3,"name": "car"}, {"supercategory": "vehicle","id": 4,"name": "motorcycle"}, {"supercategory": "vehicle","id": 5,"name": "airplane"}, {"supercategory": "vehicle","id": 6,"name": "bus"}, {"supercategory": "vehicle","id": 7,"name": "train"}, {"supercategory": "vehicle","id": 8,"name": "truck"}, {"supercategory": "vehicle","id": 9,"name": "boat"}, {"supercategory": "outdoor","id": 10,"name": "traffic light"}, {"supercategory": "outdoor","id": 11,"name": "fire hydrant"}, {"supercategory": "outdoor","id": 13,"name": "stop sign"}, {"supercategory": "outdoor","id": 14,"name": "parking meter"}, {"supercategory": "outdoor","id": 15,"name": "bench"}, {"supercategory": "animal","id": 16,"name": "bird"}, {"supercategory": "animal","id": 17,"name": "cat"}, {"supercategory": "animal","id": 18,"name": "dog"}, {"supercategory": "animal","id": 19,"name": "horse"}, {"supercategory": "animal","id": 20,"name": "sheep"}, {"supercategory": "animal","id": 21,"name": "cow"}, {"supercategory": "animal","id": 22,"name": "elephant"}, {"supercategory": "animal","id": 23,"name": "bear"}, {"supercategory": "animal","id": 24,"name": "zebra"}, {"supercategory": "animal","id": 25,"name": "giraffe"}, {"supercategory": "accessory","id": 27,"name": "backpack"}, {"supercategory": "accessory","id": 28,"name": "umbrella"}, {"supercategory": "accessory","id": 31,"name": "handbag"}, {"supercategory": "accessory","id": 32,"name": "tie"}, {"supercategory": "accessory","id": 33,"name": "suitcase"}, {"supercategory": "sports","id": 34,"name": "frisbee"}, {"supercategory": "sports","id": 35,"name": "skis"}, {"supercategory": "sports","id": 36,"name": "snowboard"}, {"supercategory": "sports","id": 37,"name": "sports ball"}, {"supercategory": "sports","id": 38,"name": "kite"}, {"supercategory": "sports","id": 39,"name": "baseball bat"}, {"supercategory": "sports","id": 40,"name": "baseball glove"}, {"supercategory": "sports","id": 41,"name": "skateboard"}, {"supercategory": "sports","id": 42,"name": "surfboard"}, {"supercategory": "sports","id": 43,"name": "tennis racket"}, {"supercategory": "kitchen","id": 44,"name": "bottle"}, {"supercategory": "kitchen","id": 46,"name": "wine glass"}, {"supercategory": "kitchen","id": 47,"name": "cup"}, {"supercategory": "kitchen","id": 48,"name": "fork"}, {"supercategory": "kitchen","id": 49,"name": "knife"}, {"supercategory": "kitchen","id": 50,"name": "spoon"}, {"supercategory": "kitchen","id": 51,"name": "bowl"}, {"supercategory": "food","id": 52,"name": "banana"}, {"supercategory": "food","id": 53,"name": "apple"}, {"supercategory": "food","id": 54,"name": "sandwich"}, {"supercategory": "food","id": 55,"name": "orange"}, {"supercategory": "food","id": 56,"name": "broccoli"}, {"supercategory": "food","id": 57,"name": "carrot"}, {"supercategory": "food","id": 58,"name": "hot dog"}, {"supercategory": "food","id": 59,"name": "pizza"}, {"supercategory": "food","id": 60,"name": "donut"}, {"supercategory": "food","id": 61,"name": "cake"}, {"supercategory": "furniture","id": 62,"name": "chair"}, {"supercategory": "furniture","id": 63,"name": "couch"}, {"supercategory": "furniture","id": 64,"name": "potted plant"}, {"supercategory": "furniture","id": 65,"name": "bed"}, {"supercategory": "furniture","id": 67,"name": "dining table"}, {"supercategory": "furniture","id": 70,"name": "toilet"}, {"supercategory": "electronic","id": 72,"name": "tv"}, {"supercategory": "electronic","id": 73,"name": "laptop"}, {"supercategory": "electronic","id": 74,"name": "mouse"}, {"supercategory": "electronic","id": 75,"name": "remote"}, {"supercategory": "electronic","id": 76,"name": "keyboard"}, {"supercategory": "electronic","id": 77,"name": "cell phone"}, {"supercategory": "appliance","id": 78,"name": "microwave"}, {"supercategory": "appliance","id": 79,"name": "oven"}, {"supercategory": "appliance","id": 80,"name": "toaster"}, {"supercategory": "appliance","id": 81,"name": "sink"}, {"supercategory": "appliance","id": 82,"name": "refrigerator"}, {"supercategory": "indoor","id": 84,"name": "book"}, {"supercategory": "indoor","id": 85,"name": "clock"}, {"supercategory": "indoor","id": 86,"name": "vase"}, {"supercategory": "indoor","id": 87,"name": "scissors"}, {"supercategory": "indoor","id": 88,"name": "teddy bear"}, {"supercategory": "indoor","id": 89,"name": "hair drier"}, {"supercategory": "indoor","id": 90,"name": "toothbrush"}],"images": [{"license": 1,"file_name": "COCO_val2014_000000581062.jpg","coco_url": "http://images.cocodataset.org/val2014/COCO_val2014_000000581062.jpg","height": 375,"width": 500,"date_captured": "2013-11-20 09:12:04","flickr_url": "http://farm4.staticflickr.com/3582/3328164554_6765a03a6a_z.jpg","id": 581062}, {"license": 5,"file_name": "COCO_val2014_000000581317.jpg","coco_url": "http://images.cocodataset.org/val2014/COCO_val2014_000000581317.jpg","height": 354,"width": 640,"date_captured": "2013-11-24 07:57:55","flickr_url": "http://farm1.staticflickr.com/14/19721466_b561e5955f_z.jpg","id": 581317}, {"license": 2,"file_name": "COCO_val2014_000000000139.jpg","coco_url": "http://images.cocodataset.org/val2014/COCO_val2014_000000000139.jpg","height": 426,"width": 640,"date_captured": "2013-11-21 01:34:01","flickr_url": "http://farm9.staticflickr.com/8035/8024364858_9c41dc1666_z.jpg","id": 139}, {"license": 3,"file_name": "COCO_val2014_000000000632.jpg","coco_url": "http://images.cocodataset.org/val2014/COCO_val2014_000000000632.jpg","height": 483,"width": 640,"date_captured": "2013-11-20 21:14:01","flickr_url": "http://farm2.staticflickr.com/1241/1243324748_eea455da9f_z.jpg","id": 632}],"annotations": [{"segmentation": [[134.12, 155.64, 136.9, 139.37, 135.31, 121.52, 133.33, 100.09, 139.68, 82.63, 131.74, 76.67, 132.93, 73.5, 149.2, 67.55, 147.61, 59.21, 154.76, 51.28, 163.88, 52.86, 167.06, 69.93, 181.34, 73.1, 194.44, 73.9, 194.44, 77.86, 183.72, 77.86, 162.3, 76.67, 157.14, 98.9, 165.07, 120.72, 162.69, 138.18, 164.28, 147.31, 171.42, 162.79, 163.49, 162.39, 154.36, 153.66, 158.33, 128.26, 148.41, 115.96, 146.03, 129.45, 142.45, 153.26, 140.87, 164.77, 133.33, 162.39]],"area": 2470.656449999998,"iscrowd": 0,"image_id": 581062,"bbox": [131.74, 51.28, 62.7, 113.49],"category_id": 1,"id": 469201}, {"segmentation": [[129.18, 160.51, 132.46, 159.07, 134.92, 163.17, 139.63, 163.58, 142.09, 158.25, 158.69, 155.18, 163.2, 159.07, 164.63, 161.53, 169.55, 160.3, 167.71, 154.57, 175.7, 152.72, 178.57, 153.34, 178.77, 157.23, 170.78, 162.15, 170.78, 166.04, 164.43, 165.84, 163.81, 164.4, 146.19, 166.66, 144.96, 169.73, 141.07, 169.93, 138.61, 167.89, 128.77, 162.15]],"area": 364.30039999999985,"iscrowd": 0,"image_id": 581062,"bbox": [128.77, 152.72, 50.0, 17.21],"category_id": 41,"id": 638773}, {"segmentation": [[240.86, 211.31, 240.16, 197.19, 236.98, 192.26, 237.34, 187.67, 245.8, 188.02, 243.33, 176.02, 250.39, 186.96, 251.8, 166.85, 255.33, 142.51, 253.21, 190.49, 261.68, 183.08, 258.86, 191.2, 260.98, 206.37, 254.63, 199.66, 252.51, 201.78, 251.8, 212.01]],"area": 531.8071000000001,"iscrowd": 0,"image_id": 139,"bbox": [236.98, 142.51, 24.7, 69.5],"category_id": 64,"id": 26547}, {"segmentation": [[9.66, 167.76, 156.35, 173.04, 153.71, 256.48, 82.56, 262.63, 7.03, 260.87]],"area": 13244.657700000002,"iscrowd": 0,"image_id": 139,"bbox": [7.03, 167.76, 149.32, 94.87],"category_id": 72,"id": 34646}, {"segmentation": [[563.33, 209.19, 637.69, 209.19, 638.56, 287.92, 557.21, 280.04]],"area": 5833.117949999999,"iscrowd": 0,"image_id": 139,"bbox": [557.21, 209.19, 81.35, 78.73],"category_id": 72,"id": 35802}, {"segmentation": [[368.16, 252.94, 383.77, 255.69, 384.69, 235.49, 389.28, 226.31, 392.03, 219.89, 413.15, 218.05, 411.31, 241.92, 411.31, 256.61, 412.23, 274.05, 414.98, 301.6, 414.98, 316.29, 412.23, 311.7, 406.72, 290.58, 405.8, 270.38, 389.28, 270.38, 381.01, 270.38, 383.77, 319.04, 377.34, 320.88, 377.34, 273.14, 358.98, 266.71, 358.98, 253.86, 370.91, 253.86]],"area": 2245.34355,"iscrowd": 0,"image_id": 139,"bbox": [358.98, 218.05, 56.0, 102.83],"category_id": 62,"id": 103487}, {"segmentation": [[319.32, 230.98, 317.41, 220.68, 296.03, 218.0, 296.03, 230.22, 297.18, 244.34, 296.03, 258.08, 299.47, 262.28, 296.03, 262.28, 298.32, 266.48, 295.27, 278.69, 291.45, 300.45, 290.69, 310.76, 292.21, 309.23, 294.89, 291.67, 298.32, 274.11, 300.61, 266.1, 323.9, 268.77, 328.09, 272.59, 326.57, 297.02, 323.9, 315.34, 327.71, 316.48, 329.62, 297.78, 329.62, 269.53, 341.84, 266.48, 345.27, 266.1, 350.23, 265.72, 343.74, 312.28, 346.8, 310.76, 351.76, 273.35, 352.52, 264.95, 350.23, 257.32, 326.19, 255.79, 323.51, 254.27, 321.61, 244.72, 320.46, 235.56, 320.08, 231.36]],"area": 1833.7840000000017,"iscrowd": 0,"image_id": 139,"bbox": [290.69, 218.0, 61.83, 98.48],"category_id": 62,"id": 104368}, {"segmentation": [[436.06, 304.37, 443.37, 300.71, 436.97, 261.4, 439.71, 245.86, 437.88, 223.92, 415.94, 223.01, 413.2, 244.95, 415.03, 258.66, 416.86, 268.72, 415.03, 289.74, 419.6, 280.6, 423.26, 263.23, 434.23, 265.98]],"area": 1289.3734500000014,"iscrowd": 0,"image_id": 139,"bbox": [413.2, 223.01, 30.17, 81.36],"category_id": 62,"id": 105328}, {"segmentation": [[317.4, 219.24, 319.8, 230.83, 338.98, 230.03, 338.58, 222.44, 338.58, 221.24, 336.58, 219.64, 328.19, 220.44]],"area": 210.14820000000023,"iscrowd": 0,"image_id": 139,"bbox": [317.4, 219.24, 21.58, 11.59],"category_id": 62,"id": 110334}, {"segmentation": [[428.19, 219.47, 430.94, 209.57, 430.39, 210.12, 421.32, 216.17, 412.8, 217.27, 413.9, 214.24, 422.42, 211.22, 429.29, 201.6, 430.67, 181.8, 430.12, 175.2, 427.09, 168.06, 426.27, 164.21, 430.94, 159.26, 440.29, 157.61, 446.06, 163.93, 448.53, 168.06, 448.53, 173.01, 449.08, 174.93, 454.03, 185.1, 455.41, 188.4, 458.43, 195.0, 460.08, 210.94, 462.28, 226.61, 460.91, 233.76, 454.31, 234.04, 460.08, 256.85, 462.56, 268.13, 465.58, 290.67, 465.85, 293.14, 463.38, 295.62, 452.66, 295.34, 448.26, 294.52, 443.59, 282.7, 446.06, 235.14, 446.34, 230.19, 438.09, 232.39, 438.09, 221.67, 434.24, 221.12, 427.09, 219.74]],"area": 2913.1103999999987,"iscrowd": 0,"image_id": 139,"bbox": [412.8, 157.61, 53.05, 138.01],"category_id": 1,"id": 230831}, {"segmentation": [[384.98, 206.58, 384.43, 199.98, 385.25, 193.66, 385.25, 190.08, 387.18, 185.13, 387.18, 182.93, 386.08, 181.01, 385.25, 178.81, 385.25, 175.79, 388.0, 172.76, 394.88, 172.21, 398.72, 173.31, 399.27, 176.06, 399.55, 183.48, 397.9, 185.68, 395.15, 188.98, 396.8, 193.38, 398.45, 194.48, 399.0, 205.75, 395.43, 207.95, 388.83, 206.03]],"area": 435.1449499999997,"iscrowd": 0,"image_id": 139,"bbox": [384.43, 172.21, 15.12, 35.74],"category_id": 1,"id": 233201}, {"segmentation": [[5.45, 269.03, 25.08, 273.39, 44.7, 289.75, 287.84, 318.09, 346.71, 321.36, 368.52, 333.36, 404.5, 419.49, 385.96, 468.55, 385.96, 475.1, 8.72, 471.82, 3.27, 266.85]],"area": 64019.87940000001,"iscrowd": 0,"image_id": 632,"bbox": [3.27, 266.85, 401.23, 208.25],"category_id": 65,"id": 315724}, {"segmentation": [[216.05, 228.68, 223.38, 228.95, 227.18, 227.46, 230.98, 218.91, 230.71, 212.94, 231.11, 207.78, 232.34, 206.43, 230.71, 204.26, 227.86, 203.71, 224.47, 204.53, 220.53, 204.26, 220.8, 202.49, 224.47, 201.54, 227.04, 201.95, 226.91, 198.97, 222.57, 196.79, 229.21, 195.98, 237.76, 195.44, 241.29, 193.0, 241.97, 190.28, 244.14, 187.98, 243.05, 185.26, 235.73, 182.28, 230.3, 182.28, 227.99, 183.09, 225.69, 183.77, 227.86, 178.21, 227.99, 165.05, 222.43, 162.6, 221.35, 158.94, 216.6, 151.89, 216.05, 150.26, 220.8, 151.21, 222.84, 152.97, 223.52, 149.04, 219.58, 147.14, 214.15, 147.41, 213.07, 142.39, 207.24, 136.69, 205.2, 136.56, 205.06, 139.13, 208.32, 144.15, 210.9, 148.09, 212.26, 149.31, 207.91, 150.53, 205.06, 153.24, 202.89, 157.18, 201.94, 160.3, 201.94, 162.47, 203.57, 165.73, 204.25, 166.13, 205.74, 165.73, 206.15, 167.49, 205.06, 169.25, 203.71, 171.69, 202.89, 174.14, 203.84, 175.63, 205.61, 176.17, 206.01, 177.66, 206.42, 186.62, 203.71, 182.55, 198.82, 183.36, 197.47, 183.91, 196.25, 179.97, 190.55, 179.84, 183.36, 185.53, 183.36, 191.1, 185.26, 195.03, 187.16, 196.12, 189.6, 196.52, 190.95, 197.2, 189.33, 201.95, 189.73, 208.87, 190.82, 210.9, 195.16, 212.4, 197.87, 212.8, 199.77, 212.53, 201.94, 211.99, 200.72, 208.19, 203.57, 208.06, 203.98, 211.31, 206.42, 214.43, 207.91, 217.96, 208.86, 223.12, 210.22, 226.78, 210.76, 227.73, 213.48, 227.86]],"area": 2464.9330500000005,"iscrowd": 0,"image_id": 632,"bbox": [183.36, 136.56, 60.78, 92.39],"category_id": 64,"id": 1610466}, {"segmentation": [[513.6, 205.75, 526.04, 206.52, 526.96, 208.82, 526.96, 221.72, 512.22, 221.72, 513.6, 205.9]],"area": 217.71919999999997,"iscrowd": 0,"image_id": 139,"bbox": [512.22, 205.75, 14.74, 15.97],"category_id": 78,"id": 1640282}, {"segmentation": [[493.1, 282.65, 493.73, 174.34, 498.72, 174.34, 504.34, 175.27, 508.09, 176.21, 511.83, 178.08, 513.39, 179.33, 513.08, 280.46, 500.28, 281.09, 494.35, 281.71]],"area": 2089.9747999999986,"iscrowd": 0,"image_id": 139,"bbox": [493.1, 174.34, 20.29, 108.31],"category_id": 82,"id": 1647285}, {"segmentation": [[611.35, 349.51, 604.77, 305.89, 613.14, 307.99, 619.11, 351.6]],"area": 338.60884999999973,"iscrowd": 0,"image_id": 139,"bbox": [604.77, 305.89, 14.34, 45.71],"category_id": 84,"id": 1648594}, {"segmentation": [[613.24, 308.24, 620.68, 308.79, 626.12, 354.68, 618.86, 352.32]],"area": 322.5935999999982,"iscrowd": 0,"image_id": 139,"bbox": [613.24, 308.24, 12.88, 46.44],"category_id": 84,"id": 1654394}, {"segmentation": [[457.3, 224.79, 455.98, 192.5, 464.55, 196.46, 461.25, 228.09]],"area": 203.6632999999997,"iscrowd": 0,"image_id": 632,"bbox": [455.98, 192.5, 8.57, 35.59],"category_id": 84,"id": 1657298}, {"segmentation": [[454.75, 252.97, 453.31, 286.9, 459.48, 286.7, 461.33, 253.38]],"area": 214.5320999999999,"iscrowd": 0,"image_id": 632,"bbox": [453.31, 252.97, 8.02, 33.93],"category_id": 84,"id": 1660743}, {"segmentation": [[444.76, 337.31, 446.01, 297.94, 450.08, 297.63, 449.76, 337.31]],"area": 179.1241999999999,"iscrowd": 0,"image_id": 632,"bbox": [444.76, 297.63, 5.32, 39.68],"category_id": 84,"id": 1660854}, {"segmentation": [[513.12, 226.56, 509.69, 196.01, 505.95, 193.21, 506.57, 191.02, 514.99, 193.83, 518.1, 227.8, 515.3, 227.18]],"area": 189.13310000000044,"iscrowd": 0,"image_id": 632,"bbox": [505.95, 191.02, 12.15, 36.78],"category_id": 84,"id": 1660921}, {"segmentation": [[494.99, 227.38, 491.56, 227.07, 487.51, 201.82, 490.94, 199.33, 494.37, 225.2]],"area": 91.54300000000008,"iscrowd": 0,"image_id": 632,"bbox": [487.51, 199.33, 7.48, 28.05],"category_id": 84,"id": 1661908}, {"segmentation": [[454.29, 121.35, 448.71, 125.07, 447.77, 136.25, 451.5, 140.9, 455.92, 143.0, 461.04, 138.34, 461.74, 128.33, 456.39, 121.12]],"area": 225.6642000000005,"iscrowd": 0,"image_id": 139,"bbox": [447.77, 121.12, 13.97, 21.88],"category_id": 85,"id": 1666628}, {"segmentation": [[553.14, 392.76, 551.33, 370.57, 549.06, 360.15, 571.7, 314.42, 574.42, 309.43, 585.74, 316.23, 579.4, 323.93, 578.04, 331.17, 577.59, 334.34, 578.95, 342.49, 581.21, 351.55, 585.29, 370.57, 584.83, 389.13, 580.31, 399.1, 563.55, 399.1, 554.95, 394.57]],"area": 2171.6188500000007,"iscrowd": 0,"image_id": 139,"bbox": [549.06, 309.43, 36.68, 89.67],"category_id": 86,"id": 1667817}, {"segmentation": [[361.37, 229.69, 361.56, 226.09, 360.42, 220.02, 359.29, 213.39, 358.72, 208.84, 353.41, 209.22, 353.22, 211.31, 353.41, 216.23, 352.65, 220.21, 350.76, 228.55, 351.71, 231.39, 360.8, 231.39, 362.13, 228.74]],"area": 178.18510000000012,"iscrowd": 0,"image_id": 139,"bbox": [350.76, 208.84, 11.37, 22.55],"category_id": 86,"id": 1669970}, {"segmentation": [[413.7, 220.47, 412.25, 231.06, 419.96, 231.54, 421.88, 219.02]],"area": 90.98724999999988,"iscrowd": 0,"image_id": 139,"bbox": [412.25, 219.02, 9.63, 12.52],"category_id": 62,"id": 1941808}, {"segmentation": [[250.96, 309.48, 251.24, 301.66, 251.52, 283.79, 247.33, 282.67, 244.82, 277.64, 248.45, 273.73, 252.36, 275.97, 260.18, 267.59, 263.81, 262.84, 269.39, 274.85, 270.51, 282.39, 273.86, 279.04, 276.93, 250.0, 276.65, 236.31, 275.81, 230.45, 304.02, 234.36, 329.71, 236.31, 327.48, 251.11, 325.52, 274.01, 325.52, 281.55, 329.99, 281.83, 333.06, 281.55, 342.28, 271.78, 348.7, 269.54, 349.54, 280.71, 346.75, 291.05, 347.86, 299.7, 347.03, 313.11, 345.91, 318.14, 334.74, 317.86, 330.55, 316.18, 331.11, 310.88, 331.67, 307.52, 331.11, 304.73, 328.31, 302.22, 271.9, 299.7, 271.07, 312.27]],"area": 5184.222599999999,"iscrowd": 0,"image_id": 632,"bbox": [244.82, 230.45, 104.72, 87.69],"category_id": 62,"id": 1947237}, {"segmentation": [[363.21, 320.43, 370.05, 332.32, 380.5, 351.05, 381.94, 355.37, 401.76, 353.21, 408.96, 350.33, 409.68, 346.37, 408.24, 344.21, 410.76, 335.92, 414.01, 318.27, 414.37, 310.34, 417.25, 308.9, 424.09, 309.26, 429.86, 292.22, 425.53, 274.93, 425.89, 254.75, 427.34, 232.18, 428.06, 218.13, 426.25, 215.61, 421.57, 225.7, 421.93, 212.37, 417.97, 212.73, 417.25, 218.49, 413.28, 217.77, 404.64, 218.49, 405.36, 222.46, 408.6, 224.98, 408.96, 228.94, 402.12, 232.55, 399.95, 237.23, 397.79, 241.55, 394.55, 243.71, 386.62, 248.4, 383.02, 249.48, 381.58, 247.32, 374.01, 246.96, 369.33, 250.92, 369.33, 251.64, 364.65, 245.51, 359.96, 247.32, 358.16, 250.56, 356.0, 254.52, 350.96, 255.6, 348.44, 259.57, 356.72, 263.53, 359.6, 266.77, 359.6, 270.37, 354.92, 271.09, 347.35, 272.18, 350.6, 275.06, 348.08, 282.98, 356.72, 287.67, 358.88, 290.55, 357.08, 300.38, 358.16, 307.59, 360.68, 312.27, 361.41, 319.47]],"area": 7034.62185,"iscrowd": 0,"image_id": 632,"bbox": [347.35, 212.37, 82.51, 143.0],"category_id": 64,"id": 1952354}, {"segmentation": [[462.16, 191.67, 488.37, 192.45, 488.68, 198.54, 488.06, 200.41, 490.24, 216.16, 490.24, 226.62, 462.01, 227.71, 463.72, 196.98, 461.38, 197.13, 460.91, 194.48]],"area": 936.6684500000003,"iscrowd": 0,"image_id": 632,"bbox": [460.91, 191.67, 29.33, 36.04],"category_id": 84,"id": 1987085}, {"segmentation": [[549.33, 248.57, 527.02, 288.31, 529.11, 289.0, 534.69, 289.0, 551.42, 254.14]],"area": 245.04669999999973,"iscrowd": 0,"image_id": 632,"bbox": [527.02, 248.57, 24.4, 40.43],"category_id": 84,"id": 1988858}, {"segmentation": [[520.64, 226.21, 519.39, 193.43, 522.21, 193.59, 523.47, 227.62]],"area": 93.38395000000153,"iscrowd": 0,"image_id": 632,"bbox": [519.39, 193.43, 4.08, 34.19],"category_id": 84,"id": 1988994}, {"segmentation": [[497.73, 55.43, 497.39, 82.79, 501.47, 82.62, 501.47, 55.43]],"area": 106.6308000000007,"iscrowd": 0,"image_id": 632,"bbox": [497.39, 55.43, 4.08, 27.36],"category_id": 84,"id": 1989419}, {"segmentation": [[524.32, 100.04, 524.99, 124.6, 525.25, 131.41, 525.12, 133.68, 525.79, 135.15, 526.85, 135.15, 527.12, 119.67, 527.12, 114.33, 527.12, 111.39, 526.85, 97.38, 524.45, 97.38]],"area": 84.2943499999996,"iscrowd": 0,"image_id": 632,"bbox": [524.32, 97.38, 2.8, 37.77],"category_id": 84,"id": 1989610}, {"segmentation": [[493.97, 156.41, 525.95, 155.72, 525.95, 161.73, 493.36, 162.08]],"area": 188.38580000000024,"iscrowd": 0,"image_id": 632,"bbox": [493.36, 155.72, 32.59, 6.36],"category_id": 84,"id": 1990447}, {"segmentation": [[501.65, 257.41, 503.21, 255.06, 454.66, 245.27, 455.06, 247.62]],"area": 117.46770000000046,"iscrowd": 0,"image_id": 632,"bbox": [454.66, 245.27, 48.55, 12.14],"category_id": 84,"id": 1990614}, {"segmentation": [[464.39, 254.8, 461.77, 286.99, 466.26, 286.99, 470.01, 253.68]],"area": 163.7676500000002,"iscrowd": 0,"image_id": 632,"bbox": [461.77, 253.68, 8.24, 33.31],"category_id": 84,"id": 2143156}, {"segmentation": [[242.95, 212.06, 241.24, 199.54, 254.32, 194.99, 255.46, 200.68, 253.18, 212.62]],"area": 189.56010000000012,"iscrowd": 0,"image_id": 139,"bbox": [241.24, 194.99, 14.22, 17.63],"category_id": 86,"id": 2146194}, {"segmentation": [[339.52, 201.72, 336.79, 216.23, 346.35, 214.69, 346.52, 199.5]],"area": 120.23200000000004,"iscrowd": 0,"image_id": 139,"bbox": [336.79, 199.5, 9.73, 16.73],"category_id": 86,"id": 2146548}, {"segmentation": [[321.21, 231.22, 387.48, 232.09, 385.95, 244.95, 362.85, 245.6, 361.32, 266.53, 361.98, 310.12, 362.41, 319.5, 353.47, 320.15, 350.2, 318.41, 350.86, 277.65, 350.64, 272.63, 351.29, 246.69, 322.96, 245.17],[411.46, 232.31, 418.87, 231.22, 415.82, 248.87, 409.93, 248.87],[436.96, 232.31, 446.77, 232.09, 446.77, 235.79, 446.12, 266.53, 445.46, 303.15, 441.54, 302.93, 440.23, 296.17, 438.92, 289.85, 438.92, 244.95, 436.09, 244.29]],"area": 2362.4897499999984,"iscrowd": 0,"image_id": 139,"bbox": [321.21, 231.22, 125.56, 88.93],"category_id": 67,"id": 2204286}, {"segmentation": [[413.66, 163.89, 406.82, 147.7, 407.96, 143.37, 410.93, 142.23, 414.12, 142.92, 418.22, 149.3, 423.69, 161.15, 421.87, 162.52, 425.97, 165.71, 427.11, 170.04, 428.25, 172.78, 428.02, 175.97, 424.6, 178.48, 420.73, 178.48, 420.04, 177.11, 424.15, 175.97, 424.83, 172.78, 423.92, 171.64, 417.99, 170.73, 417.76, 167.53, 415.71, 165.03]],"area": 311.7617500000001,"iscrowd": 0,"image_id": 581317,"bbox": [406.82, 142.23, 21.43, 36.25],"category_id": 77,"id": 327762}, {"segmentation": [[604.26, 279.15, 601.26, 288.45, 585.51, 297.91, 576.96, 302.41, 577.41, 315.61, 577.41, 321.46, 588.66, 336.02, 588.06, 338.57, 583.11, 345.62, 585.96, 354.0, 450.96, 354.0, 453.67, 343.82, 461.02, 325.67, 466.72, 317.71, 469.12, 310.81, 473.77, 299.26, 477.67, 295.81, 482.02, 280.2, 483.98, 275.55, 484.88, 271.95, 486.68, 257.09, 483.67, 259.2, 480.52, 259.2, 472.42, 250.34, 459.67, 237.14, 454.12, 227.99, 449.01, 221.08, 441.06, 208.48, 435.36, 199.33, 429.06, 195.73, 421.26, 191.07, 417.2, 187.17, 416.45, 183.72, 417.2, 179.82, 414.5, 177.57, 411.65, 175.32, 412.7, 173.52, 413.45, 171.72, 410.75, 168.87, 409.85, 163.92, 412.1, 163.17, 415.1, 162.72, 418.41, 164.67, 418.86, 167.97, 417.8, 170.07, 424.26, 171.12, 425.61, 173.82, 423.36, 176.67, 426.81, 177.12, 427.86, 176.82, 427.56, 170.82, 426.36, 170.07, 426.06, 166.62, 423.36, 164.67, 421.86, 163.32, 421.86, 161.22, 427.41, 162.42, 433.41, 161.67, 438.36, 168.42, 443.76, 177.72, 446.46, 183.27, 446.76, 187.02, 453.37, 193.78, 462.37, 206.53, 473.17, 217.48, 478.57, 219.28, 482.17, 222.28, 483.67, 223.18, 485.63, 218.08, 487.13, 214.63, 487.13, 211.18, 489.83, 204.43, 499.43, 185.67, 505.43, 176.37, 507.08, 173.37, 506.75, 169.22, 507.05, 161.71, 510.05, 159.01, 513.35, 157.66, 516.96, 157.36, 520.56, 152.41, 524.01, 150.31, 525.06, 149.86, 528.66, 150.76, 529.56, 151.06, 528.66, 149.71, 525.21, 148.96, 521.76, 148.51, 519.06, 148.36, 516.06, 147.61, 514.85, 145.51, 514.1, 142.96, 514.55, 138.31, 513.05, 136.06, 512.15, 136.36, 510.8, 135.31, 512.9, 134.11, 512.0, 131.7, 508.1, 131.7, 506.45, 129.75, 507.05, 126.3, 505.55, 124.8, 500.6, 124.95, 499.25, 123.3, 499.55, 120.0, 502.55, 112.95, 504.05, 109.5, 503.15, 105.6, 502.55, 99.14, 504.35, 85.19, 500.9, 84.29, 499.4, 82.19, 502.85, 78.14, 505.55, 77.24, 507.95, 73.49, 515.0, 67.18, 525.81, 61.63, 536.16, 58.03, 546.21, 55.93, 556.42, 57.13, 568.57, 64.63, 576.37, 71.69, 583.13, 79.49, 585.53, 82.64, 587.33, 83.54, 588.38, 85.19, 588.68, 87.44, 590.78, 88.34, 591.38, 88.64, 591.08, 92.84, 587.03, 95.69, 587.93, 96.74, 590.33, 97.19, 591.38, 103.95, 592.13, 106.05, 594.83, 106.2, 595.43, 107.7, 597.83, 106.95, 599.03, 105.3, 601.58, 105.6, 602.93, 106.8, 603.98, 107.85, 606.53, 107.4, 608.93, 108.45, 609.83, 109.65, 610.88, 111.75, 608.48, 114.0, 605.03, 114.9, 603.23, 115.5, 602.48, 117.15, 604.43, 121.95, 603.68, 127.2, 605.03, 133.96, 604.43, 141.16, 600.98, 142.06, 597.08, 143.26, 593.48, 143.56, 594.68, 147.16, 595.88, 152.26, 597.53, 155.86, 598.13, 160.36, 598.43, 162.91, 604.43, 170.42, 607.43, 176.42, 610.28, 185.27, 613.28, 192.47, 612.38, 199.53, 611.18, 207.78, 608.48, 237.19, 606.23, 263.14, 606.23, 267.8]],"area": 33684.522999999994,"iscrowd": 0,"image_id": 581317,"bbox": [409.85, 55.93, 203.43, 298.07],"category_id": 1,"id": 460897}, {"segmentation": {"counts": [201255, 6, 328, 6, 142, 8, 326, 8, 140, 10, 324, 10, 138, 12, 322, 12, 137, 12, 321, 13, 137, 12, 321, 13, 137, 12, 321, 13, 137, 12, 320, 14, 18, 6, 113, 12, 320, 14, 17, 8, 112, 13, 299, 8, 12, 14, 16, 10, 111, 14, 297, 10, 11, 14, 15, 12, 81, 6, 6, 6, 11, 15, 295, 12, 4, 20, 15, 12, 80, 8, 4, 8, 11, 14, 293, 15, 2, 21, 15, 12, 79, 10, 2, 10, 1, 23, 292, 16, 1, 22, 6, 6, 3, 12, 78, 47, 291, 40, 5, 8, 2, 12, 1, 9, 12, 7, 49, 47, 290, 41, 4, 34, 10, 9, 48, 47, 289, 42, 3, 36, 8, 11, 47, 46, 289, 43, 3, 37, 6, 13, 46, 46, 288, 43, 4, 37, 6, 13, 46, 47, 287, 44, 3, 37, 6, 13, 46, 47, 287, 44, 3, 37, 6, 13, 46, 47, 287, 44, 3, 37, 6, 13, 47, 46, 287, 44, 3, 37, 6, 13, 48, 45, 287, 43, 4, 37, 6, 14, 47, 45, 287, 42, 5, 37, 7, 13, 46, 46, 288, 41, 6, 36, 8, 12, 45, 48, 287, 41, 6, 36, 8, 12, 45, 48, 287, 41, 6, 36, 8, 13, 44, 48, 288, 40, 6, 36, 7, 14, 43, 39, 8, 2, 288, 40, 7, 35, 6, 15, 42, 9, 39, 2, 287, 41, 8, 35, 4, 16, 42, 8, 40, 2, 287, 41, 9, 34, 4, 16, 42, 8, 40, 3, 286, 41, 9, 34, 4, 16, 42, 8, 40, 4, 285, 41, 9, 34, 4, 16, 42, 8, 10, 34, 285, 41, 9, 34, 4, 16, 42, 52, 285, 41, 9, 34, 4, 17, 41, 52, 285, 40, 10, 34, 4, 18, 40, 52, 285, 40, 10, 33, 5, 19, 39, 52, 285, 40, 10, 32, 6, 14, 2, 4, 38, 52, 285, 40, 11, 29, 8, 14, 2, 4, 39, 51, 285, 40, 34, 3, 11, 14, 3, 3, 40, 50, 286, 38, 49, 14, 3, 3, 40, 50, 287, 37, 49, 15, 2, 3, 40, 50, 288, 36, 49, 15, 2, 3, 40, 49, 289, 36, 48, 16, 2, 3, 40, 50, 288, 36, 48, 16, 3, 1, 41, 50, 288, 36, 48, 16, 3, 1, 42, 49, 288, 35, 49, 17, 2, 1, 42, 48, 289, 35, 49, 17, 2, 1, 42, 49, 288, 35, 49, 17, 2, 1, 36, 1, 5, 49, 288, 35, 49, 17, 3, 1, 29, 7, 5, 49, 288, 35, 48, 18, 3, 2, 19, 16, 5, 49, 197, 10, 81, 34, 49, 19, 2, 3, 9, 25, 6, 48, 196, 12, 80, 34, 49, 19, 2, 5, 2, 30, 6, 48, 195, 14, 80, 33, 49, 19, 2, 6, 1, 30, 6, 48, 194, 16, 79, 33, 49, 19, 3, 36, 6, 48, 194, 16, 79, 33, 49, 19, 3, 36, 5, 50, 193, 16, 79, 33, 49, 20, 2, 36, 5, 50, 193, 16, 79, 33, 49, 20, 2, 36, 5, 51, 192, 16, 79, 33, 49, 20, 3, 35, 5, 51, 191, 18, 79, 32, 49, 20, 3, 35, 5, 52, 189, 20, 3, 7, 68, 32, 49, 20, 3, 34, 6, 52, 188, 22, 1, 9, 68, 31, 49, 21, 2, 34, 6, 52, 187, 34, 67, 31, 49, 21, 2, 34, 6, 53, 186, 35, 66, 31, 49, 21, 3, 33, 5, 54, 186, 35, 66, 32, 48, 21, 3, 33, 5, 54, 186, 35, 66, 32, 48, 21, 3, 33, 5, 55, 185, 35, 67, 31, 48, 22, 2, 33, 5, 55, 185, 35, 67, 31, 48, 22, 2, 33, 5, 55, 185, 35, 67, 30, 49, 22, 3, 32, 4, 55, 187, 34, 67, 31, 48, 22, 3, 32, 3, 56, 188, 33, 67, 31, 48, 22, 3, 32, 3, 56, 189, 32, 67, 31, 14, 5, 29, 23, 2, 33, 2, 57, 188, 32, 68, 30, 13, 6, 29, 23, 2, 34, 1, 39, 1, 17, 188, 32, 69, 29, 12, 6, 24, 1, 5, 23, 3, 65, 2, 4, 3, 17, 189, 31, 71, 4, 2, 21, 12, 14, 23, 22, 3, 64, 10, 17, 189, 31, 77, 21, 12, 21, 17, 22, 2, 64, 10, 17, 190, 30, 79, 19, 12, 29, 9, 22, 2, 64, 10, 17, 190, 30, 79, 19, 12, 34, 4, 22, 3, 63, 11, 16, 190, 29, 80, 19, 12, 35, 3, 22, 3, 63, 12, 14, 191, 28, 81, 19, 12, 35, 3, 22, 3, 62, 14, 12, 192, 2, 107, 19, 13, 34, 4, 22, 2, 61, 17, 8, 194, 2, 107, 19, 14, 33, 4, 22, 2, 54, 1, 4, 222, 1, 107, 19, 14, 33, 4, 22, 3, 53, 227, 1, 106, 20, 14, 33, 4, 22, 3, 53, 227, 26, 81, 20, 14, 33, 4, 22, 2, 34, 1, 20, 226, 26, 81, 20, 14, 33, 4, 58, 1, 21, 225, 26, 81, 20, 15, 32, 4, 58, 1, 21, 225, 26, 81, 20, 16, 31, 4, 58, 1, 21, 225, 26, 81, 20, 17, 30, 5, 56, 3, 20, 225, 26, 82, 19, 16, 31, 5, 19, 1, 35, 5, 19, 224, 28, 80, 20, 15, 32, 5, 54, 7, 18, 223, 29, 80, 20, 14, 33, 5, 54, 7, 18, 223, 29, 80, 20, 17, 29, 6, 54, 7, 18, 223, 29, 80, 20, 26, 20, 7, 53, 8, 17, 223, 29, 80, 20, 10, 1, 24, 11, 6, 54, 8, 17, 222, 30, 80, 20, 9, 3, 32, 1, 7, 54, 8, 16, 222, 31, 80, 19, 9, 4, 40, 54, 8, 16, 222, 31, 80, 19, 9, 4, 41, 53, 8, 16, 221, 32, 80, 19, 9, 15, 30, 53, 9, 15, 220, 33, 80, 19, 9, 18, 2, 6, 19, 53, 9, 14, 221, 33, 80, 18, 10, 27, 18, 53, 10, 13, 220, 34, 80, 18, 10, 34, 11, 53, 10, 12, 220, 35, 80, 18, 10, 4, 16, 15, 11, 52, 10, 12, 220, 35, 27, 6, 34, 6, 7, 17, 11, 4, 43, 51, 10, 12, 220, 35, 26, 8, 32, 8, 6, 17, 11, 4, 43, 51, 10, 12, 220, 36, 24, 15, 4, 6, 15, 9, 6, 17, 11, 25, 22, 51, 10, 12, 220, 36, 23, 17, 2, 8, 13, 10, 6, 17, 11, 37, 10, 51, 10, 12, 220, 36, 42, 9, 12, 10, 6, 16, 12, 37, 10, 51, 10, 12, 220, 36, 63, 31, 14, 36, 11, 50, 11, 10, 221, 36, 18, 5, 21, 8, 11, 31, 15, 36, 10, 49, 13, 8, 222, 36, 17, 35, 11, 34, 12, 36, 10, 48, 16, 4, 225, 35, 17, 35, 11, 35, 11, 36, 10, 47, 247, 34, 17, 35, 11, 36, 10, 37, 9, 46, 247, 35, 17, 35, 11, 37, 9, 37, 9, 44, 248, 36, 17, 35, 10, 38, 9, 37, 9, 42, 250, 35, 18, 35, 10, 38, 9, 37, 10, 39, 252, 35, 18, 35, 10, 38, 9, 37, 11, 36, 253, 36, 18, 35, 10, 38, 9, 38, 10, 35, 254, 36, 19, 34, 9, 39, 9, 38, 10, 24, 2, 7, 256, 36, 19, 34, 9, 39, 9, 38, 9, 25, 4, 3, 257, 37, 19, 34, 9, 39, 9, 38, 9, 25, 264, 37, 19, 33, 10, 39, 9, 38, 9, 25, 264, 37, 19, 32, 11, 39, 9, 38, 9, 25, 264, 36, 21, 10, 4, 15, 13, 39, 8, 39, 9, 24, 265, 36, 22, 8, 5, 12, 16, 39, 8, 39, 9, 23, 266, 36, 15, 13, 8, 10, 17, 39, 8, 39, 9, 21, 268, 36, 14, 14, 9, 13, 13, 39, 8, 39, 9, 20, 269, 36, 13, 16, 10, 12, 12, 39, 8, 39, 9, 18, 270, 37, 12, 18, 1, 6, 5, 10, 12, 38, 8, 39, 9, 16, 271, 38, 12, 26, 3, 12, 12, 10, 8, 18, 9, 39, 9, 14, 273, 39, 11, 27, 2, 12, 14, 6, 10, 19, 8, 39, 9, 12, 275, 39, 11, 41, 31, 18, 8, 39, 9, 11, 276, 40, 10, 41, 31, 18, 8, 39, 10, 10, 276, 40, 10, 41, 31, 17, 9, 39, 11, 8, 277, 40, 10, 41, 31, 17, 10, 38, 13, 4, 279, 39, 12, 40, 31, 16, 12, 37, 296, 39, 13, 38, 33, 14, 14, 36, 297, 37, 16, 4, 2, 4, 2, 23, 35, 12, 15, 36, 298, 37, 27, 22, 38, 4, 21, 36, 299, 37, 27, 21, 63, 12, 1, 23, 300, 36, 28, 8, 1, 10, 64, 12, 1, 22, 302, 35, 30, 4, 4, 8, 65, 12, 1, 21, 304, 11, 1, 22, 40, 4, 68, 10, 3, 19, 306, 9, 4, 5, 1, 14, 113, 8, 5, 18, 308, 5, 12, 14, 115, 4, 9, 15, 326, 14, 128, 15, 327, 12, 129, 14, 329, 10, 130, 12, 333, 6, 132, 12, 472, 10, 474, 8, 477, 4, 34074],"size": [483, 640]},"area": 20933,"iscrowd": 1,"image_id": 632,"bbox": [416, 43, 153, 303],"category_id": 84,"id": 908400000632}]
}

七、COCODataset

代码请看:datasets/coco.py

c++读取图片_四、faster-rcnn源码阅读:数据流读取相关推荐

  1. faster rcnn源码解读(四)之数据类型imdb.py和pascal_voc.py(主要是imdb和roidb数据类型的解说)

    转载自:faster rcnn源码解读(四)之数据类型imdb.py和pascal_voc.py(主要是imdb和roidb数据类型的解说) - 野孩子的专栏 - 博客频道 - CSDN.NET ht ...

  2. faster rcnn源码解读总结

    转载自:faster rcnn源码解读总结 - 野孩子的专栏 - 博客频道 - CSDN.NET http://blog.csdn.net/u010668907/article/details/519 ...

  3. Faster R-CNN源码中RPN的解析(自用)

    参考博客(一定要看前面两个) 一文看懂Faster R-CNN 详细的Faster R-CNN源码解析之RPN源码解析 关于RPN一些我的想法 rpn的中心思想就是在了anchors了,如何产生anc ...

  4. faster rcnn源码解读(六)之minibatch

    转载自:faster rcnn源码解读(六)之minibatch - 野孩子的专栏 - 博客频道 - CSDN.NET http://blog.csdn.net/u010668907/article/ ...

  5. faster rcnn源码解读(五)之layer(网络里的input-data)

    转载自:faster rcnn源码解读(五)之layer(网络里的input-data) - 野孩子的专栏 - 博客频道 - CSDN.NET http://blog.csdn.net/u010668 ...

  6. faster rcnn源码理解(二)之AnchorTargetLayer(网络中的rpn_data)

    转载自:faster rcnn源码理解(二)之AnchorTargetLayer(网络中的rpn_data) - 野孩子的专栏 - 博客频道 - CSDN.NET http://blog.csdn.n ...

  7. 【Faster R-CNN论文精度系列】从Faster R-CNN源码中,我们“学习”到了什么?

    [Faster R-CNN论文精度系列] (如下为建议阅读顺序) 1[Faster R-CNN论文精度系列]从Faster R-CNN源码中,我们"学习"到了什么? 2[Faste ...

  8. faster rcnn源码解读(三)train_faster_rcnn_alt_opt.py

    转载自:faster rcnn源码解读(三)train_faster_rcnn_alt_opt.py - 野孩子的专栏 - 博客频道 - CSDN.NET http://blog.csdn.net/u ...

  9. 基于lis3dh的简易倾角仪c源码_开源网关apisix源码阅读和最佳实践

    大家应该都接手过这种项目,前人找一个开源软件改一改,发上线. 我这里便曾经遇到过类似的问题. 随着需求的增加,各种维护人员东改改西改改,原来的开源项目被改的面目全非,再也无法和上游合并. 甚至TLS协 ...

最新文章

  1. ESP32开发板 V1.0.0 Rev1 wifi 蓝牙4MB FLASH MicroPython
  2. linux关闭涉及安全的服务,Linux中关闭不必要服务减少漏洞
  3. android+布局倾斜,android – 如何在Eclipse图形布局视图中使斜...
  4. linux当前时间 按格式化,Linux下date命令,格式化输出,时间设置
  5. 通过ajax提交到url路由
  6. 上海交大张拳石:神经网络的变量交互可解释性研究
  7. 用户登录提交前,密码加密传输
  8. 第1关:HDFS的基本操作
  9. 删除顺序表中指定范围的元素
  10. Java继承关系的类的初始化和实例化的执行顺序
  11. brew mysql要多久_mac brew 安装 mysql
  12. Hbase 二级索引
  13. VS 2015 设置为透明主题
  14. UI设计师的段位和匹配能力
  15. ASP.NET Session详细介绍
  16. 旗帜工作室2021年会总结
  17. Ubuntu查看Cuda是否全部安装成功
  18. javaSE探赜索隐之三<类与对象的爱恨情仇上>
  19. 流场可视化工程dlb-dynamicdr部署日志:阶段二:工程本地编译
  20. linux下Found a swap file by the name解决

热门文章

  1. 网络数据保障ptop_网络影响未来十大预言 宽带应用将与新媒体融合
  2. 接入高德地图第三方SDK——如何获取API Key
  3. 坚果云 linux 使用方法,树莓派4B踩坑指南 - (10)安装坚果云(暂时失败)
  4. showdown让你的网站支持Markdown和代码块highlight
  5. error An unexpected error occurred: “EPERM: operation not permitted, unlink ‘C:\\Users
  6. [日更-2019.3.31]如何下载Nexus5的LineageOS14.1(cm-14.1)系统源码并编译、刷机
  7. ListView 上下拉刷新
  8. Django——08.with语句,url标签, 解析特殊字符查verbatim
  9. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift全文翻译
  10. 关于idea无法使用搜狗输入法打出汉字的解决方案(转)