SynthText数据集
SynthText(synthetic text)其实是指用代码合成的文本图像数据，它的源码在https://github.com/ankush-me/SynthText，如果有需要你可以用它的源码来合成自己的文本训练集。我们这里使用的是用这种方法得到的官方提供数据集SynthText，这个数据集包含了80万张图像，其中融入了800万个文本。
数据格式转成适用于TextBoxes++的xml格式
有许多语言可以读取并处理mat格式文件，我在这里选用python来处理。
# 读取gt.mat数据
import scipy.io as sio
data = sio.loadmat('gt.mat')
之前我直接用string来规范文本数据，但是并不如xml包来的好管理。
import scipy.io as sio
import numpy as np
import xml.dom.minidom
import sys
import random
import os
def MatRead(matfile):
data = sio.loadmat(matfile)
train_file = open('train.txt', 'w')
test_file = open('test.txt', 'w')
for i in range(len(data['txt'][0])):
contents = []
for val in data['txt'][0][i]:
v = [x.split("\n") for x in val.strip().split(" ")]
contents.extend(sum(v, []))
print >> sys.stderr, "No.{} data".format(i)
rec = np.array(data['wordBB'][0][i], dtype=np.int32)
if len(rec.shape) == 3:
rec = rec.transpose(2,1,0)
else:
rec = rec.transpose(1,0)[np.newaxis, :]
doc = xml.dom.minidom.Document()
root = doc.createElement('annotation')
doc.appendChild(root)
print("start to process {} object".format(len(rec)))
for j in range(len(rec)):
nodeobject = doc.createElement('object')
nodecontent = doc.createElement('content')
nodecontent.appendChild(doc.createTextNode(str(contents[j])))
nodename = doc.createElement('name')
nodename.appendChild(doc.createTextNode('text'))
bndbox = {}
bndbox['x1'] = rec[j][0][0]
bndbox['y1'] = rec[j][0][1]
bndbox['x2'] = rec[j][1][0]
bndbox['y2'] = rec[j][1][1]
bndbox['x3'] = rec[j][2][0]
bndbox['y3'] = rec[j][2][1]
bndbox['x4'] = rec[j][3][0]
bndbox['y4'] = rec[j][3][1]
bndbox['xmin'] = min(bndbox['x1'], bndbox['x2'], bndbox['x3'], bndbox['x4'])
bndbox['xmax'] = max(bndbox['x1'], bndbox['x2'], bndbox['x3'], bndbox['x4'])
bndbox['ymin'] = min(bndbox['y1'], bndbox['y2'], bndbox['y3'], bndbox['y4'])
bndbox['ymax'] = max(bndbox['y1'], bndbox['y2'], bndbox['y3'], bndbox['y4'])
nodebndbox = doc.createElement('bndbox')
for k in bndbox.keys():
nodecoord = doc.createElement(k)
nodecoord.appendChild(doc.createTextNode(str(bndbox[k])))
nodebndbox.appendChild(nodecoord)
nodeobject.appendChild(nodecontent)
nodeobject.appendChild(nodename)
nodeobject.appendChild(nodebndbox)
root.appendChild(nodeobject)
filename = data['imnames'][0][i][0].replace('.jpg', '.xml')
fp = open(filename, 'w')
doc.writexml(fp, indent='\t', addindent='\t', newl='\n', encoding="utf-8")
fp.close()
rad = random.uniform(10,20)
pwd = os.getcwd()
img_path = os.path.join(pwd, data['imnames'][0][i][0])
xml_path = os.path.join(pwd, filename)
file_line = img_path + " " + xml_path + '\n'
if rad > 18:
train_file.write(file_line)
else:
test_file.write(file_line)
train_file.close()
test_file.close()
生成lmdb数据
将train.txt和test.txt移到TextBoxes_plusplus/data/text目录下，再执行./create_data.sh即可得到lmdb文件。