| With the development of science and technology,especially in recent years,the rapid rise of Internet technology,there is a variety of virtual fitting software,one of the software is based on three-dimensional scene,the effect is satisfied,but the cost is high and optional clothing is limited;The other class of software is based on the phone,the dressing process is only the replacement of the head,failing to achieve the desired effect.This paper combines a digital image processing technology related to the design of a fitting software.The basic idea of the fitting software designed in this paper is to divide the model area image into the garment area of the garment image.Therefore,the focus of this paper is to introduce the application and implementation of the algorithm in image segmentation and image matching.The image segmentation mainly includes the segmentation of the human body in the image and the segmentation of the garment area.In this paper,the main integrated edge detection segmentation method,the threshold segmentation method,the regional growth segmentation method and the gray scale projection segmentation method are proposed for the regional growth segmentation An improved algorithm to make the garment area more accurate.Image matching mainly includes clothing area size matching,location matching and skin color matching.Clothing area size matching focus is to carry out the geometric transformation of the proportion of change,according to the dresser's clothing area size to adjust the size of the model area;clothing area to match the main center of the clothing center axis and face axis alignment,the use of face recognition Method to determine the face of the central axis;skin color matching is through the color model transformation of the modelclothing area in the skin color and dresser skin color uniform.The dressing system has a high degree of automation,interactive simple,wide range of applications,but also it has a certain market value. |
PDF Full Text Request