reference
提供了转换代码,需要改写部分
https://github.com/fengzhongyouxia/TensorExpand/tree/master/TensorExpand/Object detection/Data_interface/MSCOCO/labelme data
注:命名时候尽量按照要求,按照"父_子类_num"样式去命名,否则会报错list 越界(吃了大亏,改程序手改标定。。。)
同时发现,似乎不加area 参数,只用image\annotation\category就能进行训练(所以在labelme2coco.py中,就是只转换了这三个)
coco.json说明
https://blog.csdn.net/yeyang911/article/details/78675942
https://blog.csdn.net/wc781708249/article/details/79603522#可视化
这里写一个class实现以下功能,labelme2COCO.py中 的部分代码如下:
# -*- coding:utf-8 -*-
# !/usr/bin/env python
import argparse
import json
import matplotlib.pyplot as plt
import skimage.io as io
import cv2
from labelme import utils
import numpy as np
import glob
import PIL.Image
class labelme2coco(object):
def __init__(self,labelme_json=[],save_json_path='./new.json'):
'''
:param labelme_json: 所有labelme的json文件路径组成的列表
: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) # 加载json文件
self.images.append(self.image(data,num))
for shapes in data['shapes']:
label=shapes['label'].split('_')
if label[1] not in self.label:
self.categories.append(self.categorie(label))
self.label.append(label[1])
points=shapes['points']
self.annotations.append(self.annotation(points,label,num))
self.annID+=1
def image(self,data,num):
image={}
img = utils.img_b64_to_array(data['imageData']) # 解析原图片数据
# img=io.imread(data['imagePath']) # 通过图片路径打开图片
# img = cv2.imread(data['imagePath'], 0)
height, width = img.shape[:2]
img = None
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'] = label[1]
categorie['id']=len(self.label)+1 # 0 默认为背景
categorie['name'] = label[1]
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文件时报错(不知道为什么)
# list(map(int,a[1:-1].split(','))) a=annotation['bbox'] 使用该方式转成list
annotation['bbox'] = list(map(float,self.getbbox(points)))
annotation['category_id'] = self.getcatid(label)
annotation['id'] = self.annID
return annotation
def getcatid(self,label):
for categorie in self.categories:
if label[0]==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) # 画边界线
# cv2.fillPoly(img, [np.asarray(points)], 1) # 画多边形 内部像素值为1
polygons = points
mask = self.polygons_to_mask([self.height,self.width], polygons)
return self.mask2box(mask)
def mask2box(self, mask):
'''从mask反算出其边框
mask:[h,w] 0、1组成的图片
1对应对象,只需计算1对应的行列号(左上角行列号,右下角行列号,就可以算出其边框)
'''
# np.where(mask==1)
index = np.argwhere(mask == 1)
rows = index[:, 0]
clos = index[:, 1]
# 解析左上角行列号
left_top_r = np.min(rows) # y
left_top_c = np.min(clos) # x
# 解析右下角行列号
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] 对应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) # indent=4 更加美观显示
#labelme_json=glob.glob('./test_img/*.json')
labelme_json=glob.glob('./bird.json')
# labelme_json=['./1.json']
labelme2coco(labelme_json,'./new.json')
一个更改生成:
# -*- coding:utf-8 -*-
import json
import cv2
import numpy as np
labelme_json='./bird.json'
data=json.load(open(labelme_json))
data_coco={}
# images
images=[]
image={}
file_name=data['imagePath'].split('\\')[-1] # windows \\ ;linux /
image['file_name']=file_name
image['id']=0 # 每张图片对应的id都是唯一的
# img=cv2.imread(data['imagePath'])
img=cv2.imread('./eagle.jpg')
image['height']=img.shape[0]
image['width']=img.shape[1]
img=None
images.append(image)
data_coco['images']=images
# categories
categories=[]
categorie={}
categorie['supercategory']='bird'
categorie['id']=1 # id 唯一 0 默认为背景
categorie['name']='eagle' # 波斯猫
categories.append(categorie)
'''
categorie={}
categorie['supercategory']='cat'
categorie['id']=2
categorie['name']='garden Cat' # 田园猫
categories.append(categorie)
'''
data_coco['categories']=categories
# annotations
annotations=[]
annotation={}
annotation['segmentation']=[list(np.asarray(data['shapes'][0]['points']).flatten())] # data['shapes'][0]['points']
annotation['iscrowd']=0
annotation['image_id']=image['id']
annotation['bbox']=[] # 先空着,需要反算出定位框
annotation['category_id']=1
annotation['id']=1 # 第一个对象 这个ID也不能重复,如果下一张图,id不能取1,需从1 开始往下取
annotations.append(annotation)
'''
annotation={}
annotation['segmentation']=[list(np.asarray(data['shapes'][1]['points']).flatten())]
annotation['iscrowd']=0
annotation['image_id']=image['id']
annotation['bbox']=[] # 先空着,需要反算出定位框
annotation['category_id']=2
annotation['id']=2 # 第一个对象 这个ID也不能重复,如果下一张图,id不能取1,需从1 开始往下取
annotations.append(annotation)
'''
data_coco['annotations']=annotations
# 保存json文件
json.dump(data_coco,open('./new_instances_bird.json','w'),indent=4) # indent=4 更加美观显示
增加area
from shapely.geometry import Polygon
area_ = round(poly.area,6)
annotation[‘area’] = area_
# -*- coding:utf-8 -*-
# !/usr/bin/env python
import argparse
import json
import matplotlib.pyplot as plt
import skimage.io as io
import cv2
from labelme import utils
import numpy as np
import glob
import PIL.Image
from shapely.geometry import Polygon#https://shapely.readthedocs.io/en/latest/manual.html#geometric-objects
class labelme2coco(object):
def __init__(self,labelme_json=[],save_json_path='D:/haidee/new/'):
'''
:param labelme_json: 所有labelme的json文件路径组成的列表
:param save_json_path: json保存位置
'''
self.labelme_json=labelme_json#所有的json文件
self.save_json_path=save_json_path#输出文件new.json的路径
self.images=[]
self.categories=[]
self.annotations=[]
# self.data_coco = {}
self.label=[]
self.annID=1
self.height=0
self.width=0
self.save_json()
#我的数据用label标注的名称格式是:car_car_1,car_car_2,...修改label的截断以符合自己的数据,或者按照我的命名
def data_transfer(self):
for num,json_file in enumerate(self.labelme_json):
with open(json_file,'r') as fp:
data = json.load(fp) # 加载json文件
self.images.append(self.image(data,num))
for shapes in data['shapes']:
label=shapes['label'].split('_')
if label[1] not in self.label:
self.categories.append(self.categorie(label))
self.label.append(label[1])
points=shapes['points']
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']) # 解析原图片数据
# img=io.imread(data['imagePath']) # 通过图片路径打开图片
# img = cv2.imread(data['imagePath'], 0)
height, width = img.shape[:2]
img = None
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'] = label[1]
categorie['id']=len(self.label)+1 # 0 默认为背景
categorie['name'] = label[1]
return categorie
def annotation(self,points,label,num):
annotation={}
annotation['segmentation']=[list(np.asarray(points).flatten())]
poly = Polygon(points)
annotation['iscrowd'] = 0
annotation['image_id'] = num+1
area_ = round(poly.area,6)
annotation['area'] = area_
# annotation['bbox'] = str(self.getbbox(points)) # 使用list保存json文件时报错(不知道为什么)
# list(map(int,a[1:-1].split(','))) a=annotation['bbox'] 使用该方式转成list
annotation['bbox'] = list(map(float,self.getbbox(points)))
annotation['category_id'] = self.getcatid(label)
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) # 画边界线
# cv2.fillPoly(img, [np.asarray(points)], 1) # 画多边形 内部像素值为1
polygons = points
mask = self.polygons_to_mask([self.height,self.width], polygons)
return self.mask2box(mask)
def mask2box(self, mask):
'''从mask反算出其边框
mask:[h,w] 0、1组成的图片
1对应对象,只需计算1对应的行列号(左上角行列号,右下角行列号,就可以算出其边框)
'''
# np.where(mask==1)
index = np.argwhere(mask == 1)
rows = index[:, 0]
clos = index[:, 1]
# 解析左上角行列号
left_top_r = np.min(rows) # y
left_top_c = np.min(clos) # x
# 解析右下角行列号
right_bottom_r = np.max(rows)
right_bottom_c = np.max(clos)
return [left_top_c, left_top_r, right_bottom_c-left_top_c, right_bottom_r-left_top_r] # [x1,y1,w,h] 对应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) # indent=4 更加美观显示
labelme_json=[]
for i in range(0,92):#遍历所有的labelme 标准的json,载入并转换
#ind='./test_img/val_img/s{index}.json'.format(index=i)
ind='E:/Data_Set/mask-cai/four_cls/s{index}.json'.format(index=i)
labelme_json.append(ind)
lab = labelme2coco(labelme_json,'./sileiv2_with_area.json')#调用labelme2coco类
print('Saved to :',lab.save_json_path)#保存