怎么在云服务器上搭建网站,外贸如何推广,电商网站入口,pageadmin的最新版本PPOCRLabel
使用PPOCRLabel对ocr预标注结果进行纠正由于PaddleOCR代码库十分混乱,路径经常乱调pip和代码库的代码#xff08;pip库和源码冲突#xff09;,经常报错#xff0c;因此paddleocr和ppocrlabel都是使用pip包;PPOCRLabel中使用了cv2进行图片数据的读取#xff0c;…PPOCRLabel
使用PPOCRLabel对ocr预标注结果进行纠正由于PaddleOCR代码库十分混乱,路径经常乱调pip和代码库的代码pip库和源码冲突,经常报错因此paddleocr和ppocrlabel都是使用pip包;PPOCRLabel中使用了cv2进行图片数据的读取然cv2对中文路径读取有问题经常会导致一些图片数据无法进行正常数据在次基础上把读取图片由cv2改为pilow可以规避很多读取错误
安装
pip install PPOCRLabel2.1.3启动
PPOCRLabel --lang ch修改记录
修改1
报错AttributeError: ‘NoneType’ object has no attribute ‘shape’
NoneType object has no attribute shape
Traceback (most recent call last):File D:\ProgramData\Anaconda3\envs\paddle\lib\site-packages\PPOCRLabel\libs\autoDialog.py, line 41, in runh, w, _ cv2.imdecode(np.fromfile(Imgpath, dtypenp.uint8), 1).shape
AttributeError: NoneType object has no attribute shape原因cv2不支持读取带有中文字符路径文件 h, w, _ cv2.imdecode(np.fromfile(Imgpath, dtypenp.uint8), 1).shape解决使用PIL替换cv2, 将上述代码替换为
# 修改文件D:\ProgramData\Anaconda3\envs\paddle\lib\site-packages\PPOCRLabel\libs\autoDialog.py, line 41
# 增加PIL读取图片函数
from PIL import Image
def load_image(image_path: str, return_chw: bool True, size: tuple None):image Image.open(image_path).convert(RGB)if size is not None:image image.resize(size) # resize imageimage np.asarray(image)image image[:, :, ::-1] # flip color channels from RGB to BGRw, h image.shape[1], image.shape[0] # update size after resizeif return_chw:image image.transpose(2, 0, 1)return image, (w, h)# 替换上面代码
try:image_data, (w, h) load_image(Imgpath, return_chwFalse)
except Exception as e:print(fload file {Imgpath} fail!)continue修改2
报错AttributeError: ‘NoneType’ object has no attribute ‘shape’
Traceback (most recent call last):File D:\ProgramData\Anaconda3\envs\paddle\lib\site-packages\PPOCRLabel\PPOCRLabel.py, line 1889, in saveFileself._saveFile(imgidx, modemode)File D:\ProgramData\Anaconda3\envs\paddle\lib\site-packages\PPOCRLabel\PPOCRLabel.py, line 1934, in _saveFileself.openNextImg()File D:\ProgramData\Anaconda3\envs\paddle\lib\site-packages\PPOCRLabel\PPOCRLabel.py, line 1880, in openNextImgself.loadFile(filename)File D:\ProgramData\Anaconda3\envs\paddle\lib\site-packages\PPOCRLabel\PPOCRLabel.py, line 1550, in loadFileheight, width, depth cvimg.shape
AttributeError: NoneType object has no attribute shape原因还是cv2不能读取中文路径文件解决
# 修改: File D:\ProgramData\Anaconda3\envs\paddle\lib\site-packages\PPOCRLabel\PPOCRLabel.py, line 1550, in loadFile
# cvimg cv2.imdecode(np.fromfile(unicodeFilePath, dtypenp.uint8), 1)
cvimg, _ load_image(unicodeFilePath, return_chwFalse)修改3 报错:error: (-215:Assertion failed) _src.total() 0 in function ‘cv::warpPerspective’ 报错描述在对PPOCRLABEL的框进行重新识别是发生如下报错 Can not recognise the detection box in xxxx,png. Please change manuallyunicodeFilePath is J:\data\mllm-data\xxxxxxxxx\wKh2CWERPJOAY2x-AAE62o598k0620.pngOpenCV(4.2.0) C:\projects\opencv-python\opencv\modules\imgproc\src\imgwarp.cpp:3143: error: (-215:Assertion failed) _src.total() 0 in function cv::warpPerspective原因是我们的ocr预标注数据Label.txt是使用PIL读取图片数据调用ppocr进行生产的并不是在PPOCRLabel工具内部生产的, 当我们修改数据框后, PPOCRLabel尝试再次使用cv2进行读取原图,此时由于cv2对路径较为敏感,经常会读取文件失败,才会出现如上情况 解决 依旧是修改PPOCRLabel源码, 把cv2读取改为PIL读取,就不惯着cv2的臭毛病…
# 修改如下代码def reRecognition(self):#img cv2.imdecode(np.fromfile(self.filePath,dtypenp.uint8),1)img, _ load_image(self.filePath, return_chwFalse)修改4
运行PPOCRLabel源码paddleocr使用pip安装,报错 AttributeError: ‘Namespace’ object has no attribute ‘return_word_box’ - File G:\dongyongfei786\paddle\PaddleOCR\ppstructure\predict_system.py, line 82, in __init__self.return_word_box args.return_word_box
AttributeError: Namespace object has no attribute return_word_box原因: paddleocr使用pip安装的源码中paddleocr2.7.0.3, D:\ProgramData\Anaconda3\Lib\site-packages\paddleocr\tools\infer\utility.py, 缺少 # extended functionparser.add_argument(--return_word_box, typestr2bool, defaultFalse, helpWhether return the bbox of each word (split by space) or chinese character. Only used in ppstructure for layout recovery)赞赏
都说书中自有黄金屋在这个“以钱为尊”的年代没钱那可是万万不能如果上述对各位帅哥美女有帮助的话也可动一下发财的小手你的支持的我做大的动力后续有需要考虑打成一个whl供需要的小伙伴使用后续如果还有改动会继续更新
