<![CDATA[Inaccurate clothing size selection is a common issue in online shopping, as many consumers do not know their exact body measurements. This study developed an automatic clothing size recommendation system based on image processing using MediaPipe to detect body keypoints from user images. Body parameters such as height, shoulder width, and chest circumference were calculated using the Euclidean Distance method and converted into centimeters through height-based calibration. These values were then used as input for clothing size classification using the Support Vector Machine (SVM) algorithm with a Radial Basis Function (RBF) kernel. The system was built with a MATLAB-based interface. A total of 231 body image data were used as the dataset. The classification testing results showed an accuracy of 91%, with high precision, recall, and F1-score values. Based on the evaluation, the system’s Mean Absolute Error (MAE) was 1.66 cm for body height, 1.08 cm for shoulder width, and 2.99 cm for chest circumference. The system proved to be sufficiently accurate and can assist users in automatically and efficiently determining their clothing size.]]>
Copyrights © 2025
