In this research, a new method developed with pixel coordinates is presented to obtain 3D body measurements from 2D images, to eliminate the need of the body to take measurements and to solve unfit problem in e-shopping. Front and side photographs are processed with new developed programmes named PixelList1.0 and Python 2.0. The sharp silhouettes, presented in authors’ another publication, are benefited with algorithms to reach pixel coordinates of feature points. Number of pixels in between feature points are saved, absolute values of differences consecutively are calculated, and 3D body measurements in pixels are obtained with different mathematical models. A conversion coefficient is introduced and size of a pixel is calculated accurately, 0.239 cm/pixel. 3D body measurements in centimeters are obtained by multiplying number of pixels by the size of one pixel in circle, ellipse, and hippodrome circumference equations. An overall accuracy rate of 93 % is concluded with the ellipse circumference equation.
Body Measurements 3D 2D Pixel coordinate Feature points Mathematical models
Body Measurements 3D 2D Pixel coordinate Feature points Mathematical models
Birincil Dil | İngilizce |
---|---|
Konular | Giyilebilir Malzemeler |
Bölüm | Makaleler |
Yazarlar | |
Erken Görünüm Tarihi | 1 Temmuz 2024 |
Yayımlanma Tarihi | 30 Haziran 2024 |
Gönderilme Tarihi | 8 Kasım 2022 |
Kabul Tarihi | 28 Nisan 2023 |
Yayımlandığı Sayı | Yıl 2024 Cilt: 34 Sayı: 2 |