Multi-touch object recognition: What is it? How does it work?

Have you ever wondered what object recognition is, how it works and which type is best?

In this post, we explain the different types of touchscreen object recognition in detail, plus we reveal our tips for hardware research and what to look out for - so you too can become an object recognition expert.

Object recognition - what is it?

There are different types of object recognition or object detection. Basically, the term refers to procedures for identifying known objects within an object space by means of optical, acoustic or other physical recognition procedures.

In this article, we take a closer look at object recognition in the context of multitouch applications on horizontal displays. Real objects are recognised by a touch screen. Placing a haptic object on a table is something quite common. This makes object recognition on horizontal multitouch screens a particularly intuitive process for young and old alike, in which objects can be playfully and interactively linked with digital content.

What types of object recognition are available on touchscreens?

Optical object recognition

In the old days we had optical Object Recognition for infrared cameras.

In this type of object recognition, physical objects are identified and classified according to characteristic features, such as colour, proportions or textures, using image processing. It is used, for example, in camera-based driver assistance systems to recognise traffic signs. In the context of multi-touch displays, there are also optical systems that enable tangible object recognition on the touch display surface by means of cameras integrated into the screen. As a result, these optical touch displays and tables usually have a greater overall height. Objects are equipped with an optical marker on the underside, similar to a QR code, and are then individually recognised by the system with the help of image processing and algorithms.

Advantages:

  • Easy to manage, as object markers can simply be printed
  • Works regardless of object material

Disadvantages:

  • Optical systems are very sensitive to ambient light and light changes
  • More susceptible to errors
  • Higher latency and inaccuracy in rotation angle and movement
  • Displays or tables are larger and heavier
  • Optical touch systems have poorer image quality (maximum Full HD)

Due to the noticeable disadvantages of optical touch displays, a different type of technology has prevailed.

Capacitive object recognition

You know them best from modern tablets and smartphones: capacitive touchscreens have long since become established. People use the technology every day, are familiar with the reaction times, the possible gestures and the behaviour. Capacitive touch sensors register a touch when they come into contact with a capacitively conductive object, such as a finger. This is done by means of a fine wire mesh (metal mesh) that is integrated behind the glass surface of the touch display. If the electrical field of this wire mesh is changed by contact with an object, the exact position is detected.

Nowadays, when we talk about capacitive touchscreens, we usually mean those with projected capacitive (PCAP) sensors. There are also surface capacitive touchscreens, but these only allow single-touch input and are more error-prone. PCAP touch displays enable multi-touch use with two or more fingers, they are particularly scratch-resistant, insensitive to dirt and enable excellent image quality, brightness and sharpness. In addition, they can be used regardless of lighting conditions and enable the recognition of interactive objects.

Advantages:

  • Highest sensitivity, precision and speed
  • Widely used and well-known through smartphones, tablets etc.
  • Compact, durable design
  • Top image quality (4K/UHD).
  • Functionality independent of lighting conditions

Disadvantages:

  • Doesn't work well when users wear gloves

When you hear Interactive Scape talk about multi-touch displays, we are talking about state-of-the-art premium PCAP displays - and it is on these that we offer our world-leading object recognition technology.

What are interactive objects? What are the differences between them?

Interactive objects for multitouch displays have many different names, such as markers, marker objects, tokens, tangibles or pucks. These refer to physical items of various shapes and sizes that are recognised by multitouch screens as sensory communication carriers.

There are different object recognition technologies. We have already briefly explained how objects are recognised optically. There are also clear differences among the capacitive multitouch objects.

Typically, a distinction is made between active and passive objects. Active objects send a digital identification, e.g. via Bluetooth, to the touchscreen. This means that the object is not only recognised capacitively by the sensor, but is also clearly identified by the digital ID. This makes an infinite number of IDs possible, which means that an unlimited number of individual contents and functions can be used simultaneously with the various objects on one display. The disadvantage of active objects, however, is that they have built-in batteries, are relatively heavy and large and need to be charged regularly - all in all, they have more potential sources of error.

Passive objects are distinguished without digital IDs like these, on the basis of a fixed built-in mechanism. Until recently, the number of IDs was limited by the quality of the built-in touch sensors and the size of the tokens. In most cases, a maximum of 10 different IDs were possible, and tokens had a diameter of more than 70 mm.

With our patented Scape X® object recognition, we were able to overcome the limitations of passive objects. With the help of built-in artificial intelligence, displays are able to perceive objects can be perceived much more accurately. With an object diameter of 48mm and a height of only 1.5mm, 24 IDs are already possible. With a diameter of 72mm, even more than 100 IDs are possible. In addition, completely transparent objects can be used for the first time, which hugely expands the design possibilities of digital applications.

Besides these more obvious points, there are differences that should be well researched and considered before investing in hardware. The best solution, of course, is to try it out for yourself before purchasing.

Asking these questions will help you distinguish good object recognition from very good object recognition:

  • How sensitive and fast are the position determination and rotation angle accuracy of the objects?
  • Can you assign IDs to the objects independently?
  • How many touch points and objects can the screen recognise simultaneously?
  • Are there any false touches when using object recognition, which could for example lead to annoying unwanted mouse clicks in touch applications?
  • Can you accidentally simulate an object with your fingers?
  • What happens if objects are left on the screen for a long time? Does the system lose them if they aren't being moved?
  • Will the object recognition be improved by firmware updates even after purchasing?

As a leading supplier of object recognition technology, we are always happy to help you with our know-how and experience - regardless of whether you want to use our applications, those of a market competitor or develop one yourself. We would also love to show you the world's best multi-touch object recognition in a live demonstration! Book your appointment here or contact us here with your questions or project ideas.

Subscribe to our Newsletter!
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.