CRNN+CTC OCR Relative
目录
从自然场景图片中进行文字识别,需要包括2个步骤:文字检测:解决的问题是哪里有文字,文字的范围有多少;文字识别:对定位好的文字区域进行识别,主要解决的问题是每个文字是什么,将图像中的文字区域进转化为字符信息。
1. DAR&OCR&STR
- Document Analysis& Recognition: aims at the automatic extraction information presented on paper.
- Documnet Image Enhancement; Layout Analysis; Character Recognition;
- Optical Character Recognition:
- conversion of images of handwritten or printed text into machine-encoded text
- the image may come from a scanned document, photo of a document, a scene-photo
- Scene Text Recognition: recognizing text in natural scene.
1. 基于RNN文字识别框架
- CNN+RNN+CTC(CRNN+CTC)
- CNN+Seq2Seq+Attention