The Sightguide is an electronic alternative for a guide dog for the blind. Michiel and Meindert van de Berg are experts by experience. Meindert is blind and Michiel is partially sighted. They very well know what requirements the product must meet. For the specific, technical development, they often call in the help of HAN students.

Embedded Vision Design
One of the students who is currently working on the development of the Sightguide is Bogdan Corbean, international student Embedded Systems Engineering at the HAN. He says: “For my minor in Embedded Vision Design, I am involved in improving the image processing techniques of the guide robot for the blind. For example, we want to be able to detect images at a greater distance. Think of logos of shops or a bus stop. For the latter we could develop a kind of QR-code, which the Sightguide would preferably be able to ‘see’ from about 10 metres distance. We are now testing the possibilities for this.”

Deep learning
“What I found eye-opening was how deep learning, a category within artificial intelligence, allows systems to work in the same way as the human brain. So with it, you can ‘teach’ devices to detect patterns through a similar neurological network as we humans have at our disposal.”

Newer functions
HAN lecturer Hugo Arends conducted his own research last year and participated in the development of a second prototype within the project. “As a university of applied sciences, we are always helping the company a step further with our input. With a new prototype or newer functions to show, Michiel and Meindert are stronger in the recruitment of financiers.”

Image recognition techniques
For HAN students it is an excellent opportunity to gain experience in practice. “This year we are focusing on image recognition techniques,” Hugo adds. “For our ESE-minor this is very relevant! It is also a very nice project. It helps people with little or no sight to maintain or regain their independence.”

New insights
Michiel van den Berg is pleased with the contributions from students. “Because the student groups change from time to time, we keep getting new insights,” he says. “But the complexity of the project requires a lot of familiarisation time, and putting together the various studies is time-consuming. That is why we are thinking about bringing in a commercial developer as well and leaving the further elaboration to students. It is in fact very valuable to have students of all nationalities and from various study disciplines (from Automotive and mathematical talents to Industrial Engineering and Management) shed their light on the Sightguide.”

Source: TechGelderland
Photography: Gerard Burgers