Determining the right shoe size is incredibly important, especially for kids. That’s because children wearing the wrong shoes in terms of size, form or function can lead to long-term foot problems. Shoes with the right fit will give the foot the space it needs to expand when weight is put on it, without the foot rubbing the inner edge of the shoe. It is also important that shoes are a good match with the width of a foot. The Stralsund-based Steinbeis Transfer Center for Image Processing and Information Technology in Medicine has been working on innovative solutions for the footwear industry for several years. Their latest development uses LED technology and provides an innovative measurement system for shoe retailers. A tablet is used to take a picture of the foot, analyze the image, and determine biometric measurements as well as shoe size.
Determining the biometric characteristics of feet for accurate shoe size measurements is important for adults as well. And directly converting shoe sizes from one sizing system to another tends to be very errorprone without accurate knowledge of the length and width of both feet. Since kids find it difficult to sit still during foot measurement, it is important to develop ways to do the job quickly. Shoe retailers need measurement techniques that are cost-effective, practical, reliable, and, above all, low-maintenance and user-friendly.
The new patent-pending Steinbeis solution is easy to use and very robust. To take the optical foot measurements, customers stand on an illuminated panel made of Plexiglas and trimmed with LEDs. The more LEDs lining the side of the plate, the better the illumination and contrast between the foot and the floor surface. This makes subsequent image analysis easier. A second person holds the tablet, ensuring both feet are within the viewfinder. Real-time analysis gives the user feedback in the form of visual cues, helping them to adjust the position of the tablet so an initial rough estimate of the shoe size can be displayed. Tapping the touchscreen activates the shutter producing a high-resolution image. The image is analyzed within seconds and a precise shoe size is given along with various sizing charts. The algorithm uses the length and width of both feet, and the complete measuring process can be completed in less than ten seconds, much faster than conventional measurement options.
One problem with trying to determine actual measurements within the captured image is how to avoid perspective distortion when the picture is taken. The proportions in the image change as the distance to the camera is increased. A critical success factor for good results and practical use of the system was the decision to use data to determine the orientation of the tablet in relation to the measuring space. This is aided by the gyroscopic sensor contained in most tablets. As a result, it was possible to create a new technique that takes perspective into account when creating the image. This optimizes measurements down to less than 1 mm.
The fact that most modern tablets feature high-resolution cameras, fast processors, and gyroscopic orientation sensors has opened up entirely new prospects for implementing intelligent algorithms in optical foot measurements and shoe size calculations. In addition to this new technological achievement, using flat, illuminated panels with strong light gives this new measurement technique the competitive edge. Platforms of any size and shape can be used or the panel can be recessed into the floor. Alternatively, a terminal can even be provided to take measurements without assistance. Add to this the extreme robustness of the system on account of the simple construction and longevity of the LEDs, which last up to 50,000 hours without significant reductions in luminosity. Since most retail outlets already use tablets for a variety of purposes, the outlay on maintenance and the initial investment are low. As expected, shoe retailers have shown strong interest in the new system. That’s why the design was registered along with the patent at the German patent office.
Prof. Dr. Hans-Heino Ehricke
Steinbeis Transfer Center Image Processing and Information Technology in Medicine (Stralsund)