-1

I am working in OCR android Application. Now I can take images and extract word easily using google vision API but the result is not 100% according to the angle of capturing the image. and illumination. So I tried to make some image processing techniques on image before extracting text. but I search a lot but I can't deiced what is the best image processing technique to use.(blurring,filtering) to smooth image and improve its quality. So if there is any libraries or guide line to follow up with it in this subject. How to improve image quality before extracting text

I testes this libraries for OCR operation

Tesseract

Google Mobile Vision APIs

4

1 回答 1

0

如果没有您的示例图像,很难准确地说出哪些图像处理功能可以帮助您。根据拍摄图像的环境,每张图像都可能以自己的方式有所不同,因此没有固定数量的图像处理功能可用于所有图像。如果您尝试过 Tesseract,他们在 Github 上有一篇文章,解释了改善图像以获得更好识别结果的最佳方法。我不能真正为您提供任何可以通过链接帮助您的特定图像处理功能,因为我不确切知道您的图像是什么样的。

提高识别结果的另一种方法是训练 OCR 引擎本身,使其能够更好地识别图像中的文本。Tesseract 也使您能够执行此操作:

https://github.com/tesseract-ocr/tesseract/wiki/TrainingTesseract-4.00

于 2017-01-27T15:45:13.390 回答