Image Recognition Technology In Packaging

150355-supermarket_apps li W540 100dpi

When we, as packaging professionals, hear or read about image recognition technology, the first images entering our brains are the applications in automatic packaging lines and robotic activities. Although developed for industrial purposes, smartphones and tablets equipped with cameras have pushed this technology outside the industrial production environment towards the commercialisation of consumer applications.

150355-Winewoo02 W540 100dpi

As short definition we could say, that image recognition technologies identify objects, for example the various packaging formats, labels and packaged consumer goods, by using a mobile app that accesses image recognition software. Image recognition software has the potential to transform a picture into a hyperlink to something on the Internet (for example, information, services, recipes, coupons or video). It can also be used to initiate a search with possibilities to compare similar products.

QR-codes (Quick Response)
Broadly speaking interactive packaging and the ability to connect the consumer with the company all started with QR. As visual QR-codes are easily created many a CPGC (consumer packaged goods company) included a QR-code on its packaging label. They are great space-savers as they eliminate the need to have separate printing space for additional information.

Bar-, 2D- and QR-codes all have one big disadvantage as they aren’t visually appealing and often frustrate the high-quality printed graphics of a packaging.  In contrast to QR, AR doesn’t need a printed image as the app scans the actual product, which allows the “identity” of a brand to be recognized

Bar-, 2D- and QR-codes all have one big disadvantage as they aren’t visually appealing and often frustrate the high-quality printed graphics of a packaging. In contrast to QR, IR doesn’t need a printed image as the app scans the actual product, which allows the “identity” of a brand to be recognized

However brands had to exercise a tour-de-force to convince the consumer to scan the QR-code and overcome his aversion as it turned out that photographing QR-codes is not his preferred hobby. And as success depends on the question: “What incentive does the brand have for me, as customer, to scan the code?”, the system failed as the consumer never received a proper answer.

Therefore QR became the perfect example of a failing high-tech approach to the consumer. It never was and never will be popular with the consumer. Too many steps and the necessity of too many various apps in an unfulfilling process with disappointing results failed to attract the consumer, and convoluted the brand.

150355-AR in supermarket W540 100dpi

The conclusion has to be that a brand has to be very selective in applying technology in its packaging. In other words it only can be used when the connectivity is simple and effective and guides the consumer into new, interesting or additional territories of use for the product.

Fortunately, in the market of the moment there are several intriguing technologies in development (some already in full operation) that push past the square or linear worlds of QR- and barcodes.

Image Recognition (IR)
In contrast to QR, Image Recognition doesn’t need a printed image as the app scans the actual product, which allows the “identity” of a brand to be recognized. This is an important benefit, as the graphics of a packaging aren’t compromised by printed codes.

Be aware that IR needs a platform. The success of a platform (or a cloud where all the brand’s images are stored, so that the image recognition software can recognise the scanned product and guide the consumer to the information he is expecting to receive) is its ability to show return on investment, and the brands that use apps receive detailed analytics on what, when, and where customers are scanning. In addition to the “cloud-platform”, recent technology allows for “on-device” image recognition as well. This means that brands have a choice between “scan, recognise and show digital content without connectivity” or “scan, recognise and show digital content with connectivity”.

150355-Catchoom Coca-Cola IMG_0483L W540 100dpi

IR is accessible from any web browser (mobile, pc, tablet etc) without the need for an app. This is an important step forwards, as it’s still uncertain whether the consumer is willing to download all types of apps. It makes it possible for anyone glancing at a website to scan a product by only clicking one button.

It’s clear that this unique opportunity to extend a brand into the consumer’s world needs the assistance of specialised companies. The marketing benefits are priceless, but the brand owner still has to prove to the consumer what’s in it for him.

Although the use of IR in digital marketing is increasing, each CPGC faces different technical challenges. This calls for a flexible, simple and cost effective IR solution, as continuously new content has to be created in order to maintain the interest of the consumer.

CraftAR toolbox
Catchoom of Barcelona is a cloud-based image recognition technology provider, which introduced recently CraftAR, an AR toolbox. CraftAR combines image recognition with AR content creation capabilities to enable the marketing departments of CPGC’s to quickly and easily transform AR applications into consumer marketing solutions.

Catchoom's Plandent Project

Catchoom’s Plandent Project

Through its range of advanced features, CraftAR offers streamlined tools to instantly scan thousands of objects and manage digital content like videos and 3D models once an object is recognized. Digital assets can be linked to real world objects with its drag and drop function, AR experiences can be embedded into an app with its mobile tracking Software Development Kit (SDK) and Application Programming Interface (API), and campaign performance and user interaction with content can be tracked and analysed in just a few clicks.

As its flagship product CraftAR’s image recognition service and AR experience content management system are available via SaaS (Software as a Service). The enterprise version can be integrated with corporate customer relationship management (CRM) and enterprise resource planning (ERP) platforms. It also offers white-labelling for customers via license, which can be installed on a customer’s managed servers.

As usual when you talk to hi-tech professionals, and have to absorb all the abbreviations and non-comprehensible words, you feel yourself swimming in a cloud, like the scanned image. I prefer a clear case-study to see what I factually can do with this new technology. So, here we go.

Image Recognition in Health Care
Patients often like to know more about what they are ingesting than what the product’s packaging or the prescription (formally “prescribing information” in the US; “Patient information leaflet” in Europe) provides.

Almirall, an international pharmaceutical company headquartered in Barcelona, Spain with a focus on four key therapeutic areas: respiratory, autoimmune, dermatological and gastrointestinal diseases, was looking for a solution for this problem, as it wanted to provide additional details about its products directly to its patients.

150355-Aplicació que converteix la càmera del mòbil en un lector de les caixes W540 100dpi

Due to legal restrictions, it was not possible to include a QR-code on the packaging, as the QR-code imprint should alter the lay-out of the packaging design offending regulatory requirements. Therefore, the company was faced with the challenge of providing more information to patients without modifying the packaging.

150355-Almirall Almax 500 Mg 30 Comprimidos Masticables W540 100dpi

Almirall turned to Augmented Reality technology to offer streamlined and simplified access to information for the end-user. The company’s immediate goal was to use image recognition to identify a packaging and, based on this identification provide a patient with information related to the product.

For this end, an app was created that integrated Catchoom’s cloud-based image recognition SDK, powered via CraftAR. Catchoom image recognition is designed to let users interact with real world objects in the most natural way possible.
By integrating Catchoom’s technology into the app, Almirall is able to offer patients an app with better image scanning and capture for improved object recognition.

150355-FotoFarma screenshot W540 100dpi

The mobile app, called FOTOFarma, offers a unique, real-time solution to patients that differs from conventional brochures and available information. FOTOFarma allows, through a photo of the packaging of one of Almirall’s products, access to additional information related to that product, such as patient advice, educational videos and instructions, a digital prospectus and other tools.

This case-study shows that with a serious technological application, there is the potential to nurture the relationship between the consumer and the consumer goods company by communicating important information in a way that is simple and effective for the end-user.

Catchoom, the Image Recognition and Augmented Reality specialist, powers more than 40 million devices today. It processes over 4 million image recognition requests every month and its software is trusted by Times Internet, Condé Nast, Plandent, Group Planeta amongst others, while since 2013 it has a technical collaboration with Intel. In addition Catchoom offers AR training services, application development services and support and maintenance to its customers. The company is headquartered in Barcelona/Spain and can be contacted at:

(This is a sponsored article)

3 responses to “Image Recognition Technology In Packaging

  1. Anton, very informative and helpful for the site’s readers. Respectfully, at Digimarc ( we’ll throw our hat in the ring. We’re the Barcode of Everything(TM) with imperceptible identities in print, packaging, audio and video. We eliminate the visual marring issues with QR codes; unlike IR our detection is absolute versus ‘probable’ match; we can trigger any action such as AR and again with higher accuracy; and for healthcare it’s important to be 100% accurate whereas IR can match to the wrong package based on a ‘similar.’ Our SDK and other tools are available for free.

    • Larry, thanks for your comment. As with all new technologies there are and will be some weaknesses to resolve. We have seen the, and you certainly have experienced, same weaknesses in the beginning of the implementation of the 2D barcode. I personally expect barcodes to be on packages for some more time, although it is a technology of the past. Nobody followed what Heinz Japan (I wrote about it years ago) did with the barcode. I think to much focus always has been on pricing and not on connecting with the consumer.

      • Anton, well appreciated. In our case we strove to understand the challenges in our close relationship with Datalogic and incorporated those learnings into our solution. We’re pleased that Wegmans has adopted with other significant retailers and CPGs in tow. We agree that visual barcodes, for example, will be around for many years and serve much needed functionality around the world. It will take time to transition, yet the benefits are too demonstrable to slow growth. The existing visual codes are an artifact, as at the time this visual representation was all that could be accommodated by technology. That is no longer the case today, as imaging systems can detect what the human eye cannot, hence the shift and benefits of imperceptible identities.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s