Machine vision programming: There’s an App for that!

02 April 2018

Neil Sandhu takes a look into the future of machine vision programming; and sees a more flexible, cost-effective solution appearing over the horizon.

In today’s information-rich world, there is an App for just about everything. All the hard work has been done for us – and there is no need to understand the programming that goes into creating the eye-catching user interface.

Could the same ever be possible in the field of machine vision? Will you be able to configure, commission and monitor a sensing application just by finding a ready-made ‘App’ suitable for your specific requirements?

Well, in fact, that technology is already on its way. In the not too distant future, a sensor manufacturer’s ‘App Store’ could be the go-to source of ready-developed application-specific solutions for all your programmable devices.

Machine builders, system integrators and developers will soon have the facility to access ‘Apps’ to easily set-up and configure the required software for their systems by using or adapting existing applications without having to start from scratch. 

SICK recently launched AppSpace, a concept that allows free and flexible customisation of applications on its range of programmable sensors and devices with ‘click and drop’ ease. Rather than being restricted to the available pre-developed proprietary software, AppSpace’s open software platform enables system integrators and OEMs to develop – and even share – their own tailor-made solutions.

Already developers working for SICK and for its customers have already formed a dynamic community that is collaborating and perfecting new ideas. In AppStudio developers are able to develop customer-specific applications easily, use the AppManager to import Apps into the sensor and adapt them to the task in hand. The AppPool cloud service then makes it easy for end-users to install, manage and allocate uploaded sensor Apps to programmable SICK devices anywhere in the world. Sales engineers can build configurations for customers using AppTemplates, without the need for programming skills, and integrators can quickly accomplish feasibility studies.

The first 'App'
The first application to be developed in the AppSpace programming environment is now complete, following work with a European confectionery manufacturer to create an ‘off-the-shelf’ label reading and verification solution. Developed initially for the customer’s high-speed chocolate packing line, SICK engineers perfected the all-in-one LabelChecker with the ability to read and verify text, numbers bar codes and 2D codes, as well as inspecting label design and print quality.

The result is a straightforward all-in-one solution based on the SICK InspectorP vision camera which provides an integral quality control system without the need for an additional evaluation unit.

The SICK LabelChecker’s capabilities encompass OCR and OCV in multiple regions and lines, including overlapping characters, as well as dot matrix printed and indented (peened) text. In addition, many standard types of 1D and 2D code can be read and verified including multiple codes in one image and comparison with alphanumeric print. Label design quality control includes type checking, print quality, shape, rotation, tilt and location checking and comparison with pre-learned label images with compilation of quality statistics.

LabelChecker is now available to any end user, system integrator or machine builder as a ready-to-use package with easy configuration and visualisation via the web server interface and no need for specialist programming or additional software licenses.

The future?
Traditionally, machine vision has required specialist programming expertise, huge processing power and bulky complex hardware. AppSpace offers more flexibility for vision engineers and programmers because they can customise a camera or sensor for themselves and, with full access to standard image libraries such as Halcon, adapt them for their own needs. It turns programmable devices into easily-configurable ones, even when for complex applications.

The beauty of AppSpace is that it enables applications to be developed that can then be made available off the shelf and easily adapted for a specific need.

The potential for advancement of sensor technology will eventually reach the limits of what is practical from a hardware perspective. However, the opportunities for software application developments are almost limitless.

Neil Sandhu is UK product manager – imaging, measurement, ranging & systems at SICK UK.

Contact Details and Archive...

Print this page | E-mail this page