Human dresses are different in thousands way. Human body image signals have big noise, a poor light and shade contrast and a narrow range of gray gradation distribution. The application of a traditional grads method or gray method to detect human body image edges can't obtain satisfactory results because of false detections and missed detections. According to the peculiarity of human body image, dyadic wavelet transform of cubic spline is successfully applied to detect the face and profile edges of human body image and Mallat algorithm is used in the wavelet decomposition in this paper.
To realize the automation of fashion industry measuring,designing and manufacturing, the auto-measurement of 3D size of human body is of great importance. The auto measurement system of 3D human body based on Charge Coupled Devices (CCD) and infrared sensors is presented in this paper. The system can measure the bare size of human body that excludes the effect of clothing quickly and accurately.
