Preprocessing Low Quality Handwritten Documents for OCR Models
Kunal Jaiswal, Avichal Suneja, Aman Kumar, Anany Ladha, Nidhi Mishra · International Journal for Research in Applied Science and Engineering Technology · 2023
Abstract: Handwriting recognition using OCR (Optical Character Recognition) is a transformative technology that is rapidly changing the way we interact with handwritten documents. OCR technology has traditionally been used for scanning printed text, but with advancements in machine learning and computer vision, it is now possible to recognize and digitize handwritten text as well. This has immense practical implications, as it enables handwritten notes, letters, and documents to be easily searchable, editable, and shareable in digital formats[1]. However, despite its potential, handwriting recognition using OCR is still a relatively nascenttechnology and faces several challenges. One of the main challenges is the recognition of handwriting styles that vary widely across individuals and cultures. Another challenge isinterpreting handwriting that is difficult to read, such as cursive writing or handwritten notes withsmudges or crossed out words. Additionally, OCR software is highly dependent on the qualityof the input image, making it important to optimize the lighting and capture settings for accurate results.[2] As OCR technology continues to improve, it hasthe potential to revolutionize the way we interact with handwritten documents and usher in a new era of digital transformation.