该脚本根据语料文件生成对应的图像文件,适用于文本识别等计算机视觉任务。A sentence to image generator for OCR.
new feature: 支持多线程,大幅提升脚本速度, multi-threading supported.
根目录下的fonts文件夹用于存放字体文件, font files should be put under /fonts directory.
imageset文件夹用于存放输出图像和映射表, images and mapping file will output into /imageset directory.
config/cfg.py中设置相关参数并存放语料文件, configure parameters and file paths here
(1)dict.txt是字典,字典文件应保证每行一个字符, this is dictionary, keep one charactor per line and no space at the beginning or the end of the line.
(2)sentences.txt是语料集, this is the file where you put sentences you want to draw.
(3)OUTPUT_DIR是文本图片和映射文件的输出路径, this should be the output path of images and mapping file.
(4)n_samples用于控制输出的图片总数, 注意脚本会将原始语料中的每一行进行随机切割与换行,因此最大输出图片数量不会超过处理后的语料行数。 this parameter controls the total images generated. Notice that sentences in corpus will be splitted into multiple lines randomly, so the max images generated will no more than the amount of lines of the splitted corpus.
脚本运行:python sample_generator.py
文件-类别序列映射表 mapping file
语料集: corpus where you put sentences