The technique of automatically extracting useful information from visual content, such as pictures, is called information extraction from images, sometimes referred to as image data extraction or image mining. For data extraction from images, OCR (Optical Character Recognition) technology creates editable text from scanned photos. During the survey, we found that ID cards are designed in different languages, Natural Language Processing (NLP) is being used for understanding and processing this type of data. During this survey, we focused on two languages, English, Hindi, the best result or high accuracy, 99.99% is achieved in the English language, OCR in high quality image by tesseract similarly in low quality by KerasOCR, as well as in the Hindi language tesseract performs 99.99% accuracy in high quality image similarly EasyOCR performs well in low quality image. The techniques that are used by researchers are machine learning, deep learning, SVM, YOLOv5, Faster R-CNN, text analysis and image pre-processing like segmentation etc. for data extraction.