Handwritten Character Recognition using Neural Network

Sahu, Manish; Dewangan, Naveen Kumar

doi:10.17148/ijarcce.2017.6603

Cited by 3 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Handwriting recognition has been a subject of research for several decades. Offline handwriting recognition automatically converts text in an image into letter codes usable within computer and text-processing applications (Sahu, 2013). Despite challenges like varying handwriting styles and image quality, advancements in machine learning have improved the accuracy of recognizing complete words and sentences.…”

Section: Introductionmentioning

confidence: 99%

Offline Handwritten Text Extraction and Recognition Using CNN-BLSTM-CTC Network

Shrestha,

Shakya

et al. 2023

Int. J. Engin. Technol.

View full text Add to dashboard Cite

Offline handwriting recognition is a significant research area that aims at tackling problems encountered with handwritten forms in college application and registration processes. The objective of this study is to address the problems of English language offline handwriting recognition via CNN-BLSTM-CTC neural network applied for an NCE Admission form. The system uses OpenCV for image processing, TensorFlow for neural network training and handwritten text recognition, and trains and tests it on the IAM database using image segmentation-based handwriting recognition. With the help of proper image verification, the system allows the users to upload images of the NCE Admission form provided that they strictly comply with the specified format; it denies access to images not conforming to the set standards. Following the successful delivery of a valid image, the form goes through extensive processing that includes text extraction from specific regions of interest (ROIs). The extracted texts are then passed to text recognition block. The recognized texts are then recorded in a CSV file under respective fields. The text recognition model has a CER of approximately 9.33%. The study performed with 15 NCE Admission forms found that the average Character Error Rate (CER) was approximately 12.2% for scanned images and 19.3% for camera-captured images. The results show that accuracy depends on aspects such as the quality and orientation of the image; thus, scanned images are preferred for better performance.

show abstract

Section: Introductionmentioning

confidence: 99%

Offline Handwritten Text Extraction and Recognition Using CNN-BLSTM-CTC Network

Shrestha,

Shakya

et al. 2023

Int. J. Engin. Technol.

View full text Add to dashboard Cite

show abstract

“…Sehingga dibutuhkan suatu support-system yang dapat membantu para investor pemula untuk membuat keputusan dalam jual-beli saham.Penelitian ini merupakan tahap pertama dari keseluruhan sistem analisa jual beli saham otomatis berbasis Optical Character Recognition (OCR) [1]. OCR sebagai bagian dari image processing adalah teknik menerjemahkan karakter teks yang ditulis tangan, diketik, atau dicetak menjadi teks yang dikodekan mesin [2]. Penerapan OCR pada penelitian ini adalah untuk mengambil citra data harga saham pada aplikasi BEI kemudian mengkonversinya menjadi data berupa text-editable.…”

unclassified

Pengenalan Pola Berbasis OCR untuk Pengambilan Data Bursa Saham

Okta

Aulia

Burhanuddin

2021

JRE

View full text Add to dashboard Cite

The investor must be able to use instinct to evaluate when to sell and buy stocks. This is, of fact, a weakness for inexperienced investors, in addition to the decision's inaccuracy and the time it takes to evaluate a slew of ineffective results. So that, a support system is needed to help the investors make decisions in buying and selling shares. This support system creates an online analysis curve display through text data in the BEI stock price application. The data processing based on pattern recognition will be carried out so that a buying and selling decision can be made to calculate the profit and loss by investors. As the first step of the whole system, this research has built an image-to-text conversion system based on OCR (Optical Character Recognition) that can convert the non-editable text (.jpg) to be editable (.text) online. After obtaining this .text data, the will used the system in further research to analyze stock buying and selling decisions. According to research on eight companies, the OCR-based image to text conversion has a 96.8% accuracy rate. Meanwhile, using Droid serif, Takao PGhotic, and Waree fonts at 12pt font sizes, it has 100 percent accuracy in Libre Office.

show abstract