YCG09 / chinese_ocr

CTPN + DenseNet + CTC based end-to-end Chinese OCR implemented using tensorflow and keras
Apache License 2.0
2.74k stars 1.08k forks source link

split_label问题 #245

Closed cellphonef closed 5 years ago

cellphonef commented 5 years ago

with open(os.path.join('label_tmp', stem) + '.txt', 'a') as f: for i in range(len(x_left)):

f.writelines("text\t")

            f.writelines(str(int(x_left[i])))
            f.writelines("\t")
            f.writelines(str(int(ymin)))
            f.writelines("\t")
            f.writelines(str(int(x_right[i])))
            f.writelines("\t")
            f.writelines(str(int(ymax)))
            f.writelines("\n")

作者改了原作者的代码把? 每行多出来一个text字符串 并且每个坐标值不是以逗号分割的而是以/t 我是要以原作者以逗号分割还是制表符分隔啊? 为什么多一行text啊 有何意义? @YCG09

cellphonef commented 5 years ago

已解决