A quick introduction to OCR and its applications

What is OCR?

"OCR or Optical Character Recognition is the recognition of text from printed or handwritten documents and images in order to distinguish alphanumeric characters using technology." ‍

That's the technical definition.

Let's look at a more practical definition.

What do you see in the below image?

Most likely, you would see the capitalized English character "A". Your mind has already done some preprocessing for you to identify light and dark regions, strokes and other features such as the triangle in the middle surrounded by darker regions.

However, this is what the computer sees when it sees the same image.

A computer simply 'sees' 1s and 0s. It has no cognition of what the patterns of ones and zeros represents to humans. OCR is the technology that converts the pattern of ones and zeros to machine readable data (eg. ASCII, HTML, JSON).

OCR technology helps computers understand printed and handwritten information by converting it to machine readable data.

How does OCR technology work?

OCR technology has come a long way since 1990s. Lets take an example. Suppose you're an OCR computer program presented with lots of different letters written in lots of different fonts; how do you pick out all the letter As if they all look slightly different?

You could use a rule like this: If you see two angled lines that meet in a point at the top, in the center, and there's a horizontal line between them about halfway down, that's a letter A.

Apply that rule and you'll recognize most capital letter As, no matter what font they're written in. Instead of recognizing the complete pattern of an A, you're detecting the individual component features (angled lines, crossed lines, or whatever) from which the character is made.

Most modern OCR programs work by feature detection. However, rather than creating specific rules for each letter, they use neural networks for it. How neural networks work is much more complicated and out of scope of this article. In short, neural networks automatically detect features provided that it is trained on a large number of samples of the character it is trying to detect.

OCR software tries to recognize characters in the image /document by slicing the image into smaller pieces and then passing each piece through a neural network to check if it contains a character and to find closest matching character. Modern OCR programs such as Google Vision and Tesseract then combine these characters based on the spacing between them to give word representations.

What are the applications of OCR?

It is quite likely that you have used OCR technology in your life if you have used an app such as CamScanner to take photos of business cards. When you upload photos & PDF files to Google Drive, Google automatically scans them using OCR technology to identify text in them. Other applications of OCR are:

• Extracting data from business documents, for example, bank checks, invoice, bank statement and receipts
• Recognizing number plate recognition in traffic cameras & CCTVs
• Extracting data from passports at airports
• Extracting data from business cards
• Key value pair and table extraction from insurance documents
• Making physical books readable online
• Making documents searchable

What are the shortcomings of OCR & where is OCR technology headed?

There are 2 main shortcomings of OCR technology: accuracy and text categorization.  

1. Accuracy

One of the issues with OCR technology is that the accuracy may not be 100%. For example in the image below "21.08.2018" could be captured as "2I.O8.2OI8". Hence, you need a second system that validates the output of the OCR engine.

‍

2. Text Categorization

OCR technology identifies characters and then combines those characters into words. However, for business use, it is important to identify what those words mean. For example, OCR technology will give the output “Invoice No: 12345” where “Invoice No” represents the “invoice_number_key” and “12345” represents “invoice_number_value". This is where you need intelligence built on top of base OCR technology to make the identified text usable.

At Docsumo, we solve both these issues. Docsumo automates data extraction from documents and makes the data actionable. Using advanced computer vision and natural language processing, it validates the extracted data so that it can be directly consumed by downstream software.

Some Commonly Asked Questions about OCR Technology

Below we answer a list of common OCR FAQs for those who want to know more about this technology:-

1. What is the full form of OCR? 

OCR stands for Optical Character Recognition. 

It is the technology used for scanning numbers, letters, shapes, and images from all sorts of documents. It is capable of reading special characters, symbols, and paragraphs from PDFs, spreadsheets, and various electronic files as well.

OCR’s history traces back to the 1920s when physicist Emanuel Goldberg created a machine that became capable of reading characters and converting them into telegraphic codes. The evolution of modern OCR took place after the 19th century when neural networks and the field of Natural Language Processing (NLP) made advancements in technological innovations.‍

2. What’s the definition of OCR scanning?

The definition of OCR scanning is to recognize and read information from physical documents, scan it, and process the data into electronic formats. The meaning of ‘scanning,’ is to retrieve information from documents and process it in file management systems.

OCR programs store information as editable text or as documents on computers. For example, if you scan a piece of paper, OCR technology will enable you to extract data from this scanned image. This process involves converting the characters from these images into a machine-readable format.

3. How does OCR scanning/processing work? 

OCR software programs let computers recognize text from physical documents, clean it up, and make it easier to interpret. OCR algorithm preprocess images from these documents and prepare them for reading in order for better chances of recognition. Common OCR scanning techniques include character isolation, aspect ratio scaling and normalization, de-skewing documents, and converting images to black and white photos for distinguishing text.. 

Zonal OCR is a subset of OCR technology which lets users scan specific “zones” or regions of documents and ignore the rest. This is useful for identifying the key-value pairs and line-items in a document, and save it instead of processing entire documents. ‍

4. What’s the use of OCR? 

OCR technology is used by different industry verticals for the purpose of scanning, storing, processing, and sharing documents. Banks do data capture and extraction using OCR algorithms to archive client-related paperwork and make digital content more accessible. Signature recognition and validation using OCR is used for detecting fraud in documents and identifying cases of forgery for processing loan applications.

The logistics industry deals with huge volumes of data and requires authorities to identify inaccuracies in documentation. OCR solutions make it easy to automate process workflows, capture and validate information, and forward alerts as EDIs to stakeholders. The logistics industry is going paperless and OCR software lets employees save time and make remote work possible by removing the need for their physical presence when it comes to submitting relevant documentation.

Real estate industry uses OCR to get faster and accurate data analysis for verifying properties. Robotic Process Automation technology embedded with modern OCR solutions helps companies save the total cost of ownership and process more than 50 million documents a year, thus drastically increasing efficiency and generating savings in sales due to genuine paperwork. Commercial real estate deals automate the underwriting process and extract data from rent rolls for faster processing using modern OCR solutions.

OCR is used by insurance companies for filing claims, performing customer profile analysis, and automating data capture to save time and reduce errors associated with manual data entry. It can take over a 100 employees to process 10,000 documents a month but OCR in insurance can finish document processing in a matter of days!

5. What are the benefits of OCR over manual data extraction?

Automated data extraction via OCR helps businesses in cutting costs and being more efficient in document processing. OCR offers users the following benefits over manual data entry:‍

1. 99% Data Accuracy – Manual data entry is laden with mistakes due to human-error and gets details often overlooked. Automated OCR solutions ensure up to 99% data accuracy, are precise, and do not misinterpret or miss details. ‍

2. Easy Document Management – Intelligent OCR solutions read information well and make it easy to store. Document management becomes convenient for businesses as they digitize files and save in document processing system‍

3. Quicker Data Processing - OCR technology is 10x faster than manual data entry and greatly improves the speed of conversions from scanning to digitizing documents‍

4. Reduced Long-Term Costs – It is agreed that the initial costs of investing in intelligent OCR solutions is high. But the long-term costs of using these programs are low since they don’t require much maintenance. ‍

5. Improved Customer Service - Customers often want quick and easily accessible ways of  systematically storing and retrieving documents at blazing fast speeds. OCR makes customer onboarding seamless for businesses and drastically improves their experience online‍

6. How accurate is OCR technology?

OCR technology is generally accepted to be 98% to 99% accurate when it comes to reading and interpreting information correctly from documents. This means that for a 1,000-page document, up to 980 or 990 characters are accurately read by the software and recorded electronically. 

Reliable OCR solutions like Docsumo don’t just have 99%+ page-level ocr accuracy but a high level of field-accuracy as well. High field-level accuracy scores let users achieve true automation when it comes to intelligent data entry and these programs require minimal manual review after data is entered by software algorithms. 

7. What are some popular OCR APIs? 

OCR APIs are designed to transcribe text from handwritten documents for interpretation by machines. Popular use cases of OCR APIs include banking, finance, legal sectors, educational institutions, and the real estate industry. For legal documents, you can use OCR APIs to transcribe documents such as affidavits, judgments, filings, etc.

OCR APIs are used for automatically processing invoices, receipts, bill of laden, and extract information from tax records. There are APIs dedicated to scanning KYC documents, survey forms, and classifying text from a variety of documents.

The most popular OCR APIs in the industry are: 

• Google Document AI
• Amazon Textract
• Abby Flexicapture
• Rossum AI
• Docsumo

8. What are the limitations of OCR?

OCR is used for extracting text data from images and classifying it using intelligent analysis. However, even OCR has a set number of limitations which are as follows:

• OCR may not correctly scan tilted text and misinterpret handwritten fonts, unlike ICR. This can make certain words or phrases undiscoverable by document processing systems
• OCR solutions may fail to interpret text contained in images. These solutions tend to partially read text from graphics and not convert the images into full text for interpretation
• Global spell checking errors and redundancy in error rates is another challenge faced by OCR software in the industry
• Incorrect document boundaries across multiple files are a classic limitation of OCR. Embedded documents may be left out and there is lack of visual classification faced for these documents‍

9. What’s the difference between OCR and ICR? 

OCR and ICR each have their own use-cases when it comes to document processing and scanning. The key difference between the two is the way data is read from paper-based documents. OCR is best used for scanning text-based documents and converting them into digital files. There is no need to manually retype data when you use OCR software and it is considered to be a very cost-effective solution for businesses.

ICR, on the other hand, is ideal for reading handwritten fonts and different styles of cursive text. It can recognize and convert multiple styles of handwriting effectively and is powered by intelligent neural networks which are capable of automatically updating databases.

Although it is more expensive than OCR, it can save countless hours of time since it virtually reads any font and prevents human input errors associated with handwritten data entry.

10. Where can I see OCR in action?

If you’re new to the world of OCR and want to give a test drive, the best way to get started is by using the Docsumo free online OCR scanner. If you have a few document samples ready, you can upload your PDFs and image files to extract data automatically. You don’t need to install the software in your system. It’s completely free to use and there are no usage limitations.

Another free OCR tool we recommend to see the technology in action is the Docsumo OCR Chrome Extension. You can use it to scan text from websites, blogs, news articles, forums, and a variety of online portals. The data read can be translated into different languages like Spanish, Portuguese, German, etc. as well at no additional cost. Docsumo’s OCR Chrome Extension is also capable of reading text from visuals, graphical elements, video thumbnails, and a variety of images online.

Docsumo uses proprietary machine learning algorithms and AI technology for automating data capture in businesses and enterprises. Besides enjoying complete data privacy and legal compliance, you can use our intelligent OCR tools to automate document processing and improve productivity at work.

To get a first-hand experience of how intelligent OCR can benefit your business, sign up for a free demo with Docsumo and experience the difference today!