QR Code

The abbreviation QR code stands for Quick Response Code. It is a two-dimensional matrix or barcode. It was developed in 1994 by the Japanese company Denso Wave. It was originally used for logistics processes in automobile production. However, due to its fast readability and high storage capacity, the QR code is now also used in many other industries. In the retail industry, for example, it is used to direct customers to websites, in marketing campaigns to improve customer interaction, or in event organization to quickly check tickets.

A QR code consists of black and white squares arranged in a square grid. These squares represent the data contained in the code. The QR code can be read with an optical reader, e.g. the camera of a smartphone. The data in a QR code can contain text, URLs, or other data such as contact information or geographical coordinates.

QR codes have several advantages over conventional barcodes. They can be read from any direction, store more data and are still readable even if parts of the code are damaged. In addition, QR codes can be configured to contain dynamic information, i.e. the data they contain can be updated without having to reprint the code itself.

The term optical identification technologies stands for technologies that are based on the transmission of optical waves instead of radio waves. It is a font and symbol-based process. Optical readability is used for automatic identification, and is therefore considered a part of Auto-ID technologies.

Optical identification technologies include barcodes, data matrix codes (2D), optical character recognition (OCR), image processing systems (cameras), and biometric identification. Essentially, optical identification enables visual information to be read and interpreted using optical sensors and special software algorithms.

The expansion of e-commerce and logistics in particular have contributed to the strong growth of optical identification technologies. At more than 35.8 percent, the warehousing sector saw the strongest growth in 2022.

In warehousing, barcodes are primarily used for warehouse management (warehouse management systems, or WMS) and order processing, but also for track and traceability. Overall, the rise in industrial production and the increasing introduction of automatic recognition systems are also driving growth in the ORM market.

The banking, financial services, and insurance (BFSI) sector has a 19 percent share of the global market. This is due to the optimization and automation of document-intensive processes.

Barcodes are read with mobile scanners, at scanning stations, and on camera-based machines or imagers, and are processed electronically. Like QR codes, they can also be read with a smartphone. High-performance image processing systems are used in the field of machine learning and artificial intelligence (AI in companies) in the context of autonomous driving.

Desktop, mobile, and industrial printers are used in the printer sector. Industrial printers will account for the largest share of revenue in 2022 at over 63.5 percent. In terms of technology, the market is divided into thermal transfer, direct transfer, laser, impact, and inkjet technology. With a revenue share of over 40.7 percent in 2022, the thermal transfer segment proves to be the dominant technology.

What connects them is the fact that both codes are two-dimensional (2D) barcodes, yet they are not the same. Both codes can store data both horizontally and vertically. This means that they contain more information than a conventional one-dimensional barcode.

In addition, both codes can be read by special scanners or smartphone cameras. A data matrix code is therefore not a QR code, but a separate type of 2D barcode.

The QR code, however, consists of black squares in a square grid on a white background. Typically, it has three distinctive larger squares in the corners of the code that serve as alignment markers. It can store up to 7,089 numeric characters or 4,296 alphanumeric characters. The QR code uses the Reed-Solomon error correction algorithm for error correction. In terms of distribution, the QR code is mainly used in the consumer sector, for marketing applications on product packaging or on tickets.

The data matrix code has a similar square grid, however, it uses a different marking system. Normally, this marking system has fixed edge lines (L-shaped) and other characteristic alignment patterns. Compared to the QR code, the data matrix code has a higher data capacity per unit area, and can store up to 2,335 alphanumeric characters.

Data matrix codes support different levels of error correction, usually depending on the application. It is more commonly used in industrial and medical applications as it can store a higher amount of data in a smaller space. This is particularly useful in the electronics and pharmaceutical industries where space is often limited.

According to a report by market research and consulting firm “Grand View Research”, the global market for optical identification technologies is set to grow by almost 15 percent between 2023 and 2030. According to a report by the British market research company “The Business Research Company”, the global market for 2D barcode readers will grow by over 7 percent in the same period.

In terms of product types, the market for handheld readers is divided into handheld readers and stationary readers. Handheld readers accounted for the largest market share of around 71.5 percent in 2022.

In particular, the growth of e-commerce and digitalization in the retail industry as well as digitalization in the healthcare, have the greatest influence on the growth of optical identification technologies, alongside the transport and warehousing industry.

Application areas include authentication solutions, track & trace, supply chain management, warehouse management, and machine vision. Machine vision is a technology that also reads visual information. However, it is not limited to codes or texts, and can also recognize entire objects, patterns, and movements.

The first development of a barcode as a binary code to simplify payment at the checkout took place as early as 1948 in the USA. The patent for this innovation was then registered in the USA in 1952.

In the 1970s, the barcode was further developed into the Universal Product Code (UPC). This was followed in 1973 by the development of the Code 39 barcode symbology. A milestone in 1974 was the first successful scanning of a barcode on a chewing gum pack in the USA. At the same time, the EAN system (European Article Number) was introduced in Europe in 1976.

The 1980s marked a further development phase in which the Data Matrix code (D2) was created. This code had the remarkable ability to encode 1556 bytes, which represented a significant increase in capacity compared to the earlier barcodes.

The QR code was first introduced by the company Denso Wave in 1994. This new code type could already encode 2953 bytes and thus opened up a wide range of applications.

Another innovation followed in 2007 with the introduction of the 4D code. Here, the fourth dimension, time, was integrated into the coding, which opened up new application and usage possibilities.

Barcode symbologies can be divided into two main categories: Linear or one-dimensional symbologies, which include UPC, Code 128, Code 39, and Interleaved 2 of 5 Code, as well as Data Matrix. Two-dimensional symbologies include the Data Matrix and the QR Code. The further development from 1D to 2D barcodes has been successful due to the introduction of information layers.

Biometric recognition methods are based on acoustic or optical processes. Optical methods include fingerprint, hand geometry, vein scan, iris, and facial recognition. Acoustic methods include voice recognition.

Optical Character Recognition (OCR) is a visual or optical technology that makes it possible to convert texts and images into editable files. It is also one of the optical identification technologies. OCR technology is able to convert graphic products in the form of text and images into separate images. By analyzing visual features such as shapes, lines, and pixel patterns, OCR systems can recognize letters, numbers, and other characters in the visual data, and convert them into machine-readable text. It is therefore also referred to as plain text reading technology.

This process involves the recognition of structures, whereby text blocks are first distinguished from graphic elements. This is followed by the recognition of line structures and the separation of individual characters. The decision as to which text character is involved is made by certain algorithms that also take the linguistic context into account.

Machine learning plays a crucial role in the further development of OCR systems. By using machine learning algorithms, these systems can be trained to extract text from images or documents even more accurately and reliably.

Pattern recognition enables the interpretation and understanding of different fonts, font sizes, and styles. Another area in which machine learning is used concerns correction mechanisms. Here, automatic error detection and correction algorithms are developed to reduce errors in text recognition, which can be caused by blurring in the image or irregular writing, for example.

In addition, adaptive training enables continuous improvement of the OCR systems. Through regular training with new data, they can adapt to different writing styles and languages and improve their accuracy.