Department of Image Processing

When Henry Ford invented the moving assembly line, he changed the production paradigm from expansive but personalized crafted items to cheap, non-customizable, mass-produced ones. Nowadays, with the progress of 3D printing, a new opportunity for the personalization of products emerged.

Posedla s.r.o. is a startup that develops personalized bicycle saddles based on customer data. The critical part of the saddle is 3D printed using thermoplastic polyurethane; hence it offers a high amount of customizability. Naturally, the personalization of this component has many advantages, mainly health benefits and increased comfort. One of the essential parameters is the width of the sitting bones, which determines where are the pressure zones on the saddle. However, acquiring this information through precise scanning might be uncomfortable for most people and require them to travel to such a machine. Instead, Posedla sends memory foam to the customer, who can easily imprint their bottom into the foam. We developed a semi-automatic method for efficient sitting bones width estimation based on customer photographs of the foam. It does not require any additional measurement or marking by the customer.