Optical character recognition (OCR) is the process of digitizing and recognizing images of written, typed or printed text [78]. A typical example of OCR usage is postal automation, where letters and packages are classified based on their hand‐written or typed addresses and labels. Rather than manually reading and typing the address of each letter into a computer, OCR allows the computer to recognize and process the addresses automatically. Other applications of OCR include capturing licence plate numbers, conversion of physical documents to editable text documents and assistance for the blind and visually impaired. A popular information source used for OCR is the MNIST dataset [79]. It consists of 70 000 greyscale images of handwritten numerical digits ranging from 0 to 9. Each digit is stored as a pixel image. The dataset is divided into 60 000 training samples and 10 000 test samples. Figure 8.1a displays a few examples from the MNIST dataset [80]. The MNIST website contains the performance of various classification methods trained on the dataset. It is reported on the MNIST website that CNNs currently outperform every other classifier for character recognition applications. In Figure 8.1b, the lowest error rate corresponding to each classifier is listed.
Figure 8.1 Essential information for OCR applications in MNIST ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month, and much more.
O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.
Julian F.
Head of Cybersecurity
I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.
Addison B.
Field Engineer
I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.
Amir M.
Data Platform Tech Lead
I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.
O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.
pixel image. The dataset is divided into 60 000 training samples and 10 000 test samples. Figure 8.1a displays a few examples from the MNIST dataset [80]. The MNIST website contains the performance of various classification methods trained on the dataset. It is reported on the MNIST website that CNNs currently outperform every other classifier for character recognition applications. In Figure 8.1b, the lowest error rate corresponding to each classifier is listed.