Developing an Anthropometric Measurement System with a Hybrid Model (CNN+RF) integrated on the Android Studio Platform

Abstract

Nowadays, cosmetic surgery is becoming a rapidly growing industry in many countries. Identifying landmarks and measuring anthropometric dimensions on the human face are essential for determining balanced and harmonious facial features, offering valuable insights for cosmetic surgery, a highly relevant field today. However, the process of marking and measuring facial anthropometric data is time-consuming, costly, and often performed by specialists. In this study, a cost-effective anthropometric measurement system is proposed, which can be utilized on an Android platform via mobile devices, enabling individuals to easily measure their own anthropometric data. This system serves as a foundation for integration into medical devices to assist doctors in facial surgery for patients. A hybrid model is applied to detect facial landmarks and calculates anthropometric measurements based on medical theories. Landmarks are learned and identified by this model through feature extraction during training. In this study, we will use hybrid machine learning methods, including Convolutional Neural Networks (CNN) and Random Forest (RF), which will be incorporated into an application created with Android Studio. The result is an interactive application that provides personalized and harmonious anthropometric facial data tailored to the needs of cosmetic surgery. The development and application of this technology is expected to bring numerous benefits, including reducing costs and time for both patients and doctors, while also enhancing the efficiency and accuracy of cosmetic surgery procedures