Custom Made Cycling Jerseys Prediction Based on Kinect Analysis for Improved Performance

Abstract

Human factors of cycling jerseys allow supporting the performance of cyclists in terms of aerodynamics, biomechanics and physical comfort. Within this research, it is aimed to evaluate three contactless methods for predicting body measurements that allows selecting the size of a cycling jersey. The accuracy of 2D images, 3D markers and a 3D scan technique are compared to hand measurements. With respect to shoulder width, RSME is 2.8 cm for 2D images, 15.1 cm for markers and 8.5 cm for the full body scanner. The results suggest that 2D images may be a useful, low-cost and accurate method for predicting body size measurements of cycling clothing. A careful selection of body sizes or a combination thereof, can aid to enhance the accuracy of a contactless body size prediction for selecting the appropriate cycling jersey size.