FOOT AI PREDICTION: AN ANDROID-BASED ARTIFICIAL INTELLIGENCE APPLICATION FOR AUTOMATED AGE AND GENDER ESTIMATION FROM DIGITAL BAREFOOT PRINTS

Abstract

Rapid and reliable demographic profiling is essential in forensic investigations, particularly in situations where conventional biometric evidence, such as fingerprints or DNA, is unavailable or unsuitable for identification. Barefoot prints preserve distinctive morphological features that can provide valuable information for estimating demographic characteristics. However, conventional footprint analysis remains dependent on manual anthropometric measurements, expert interpretation, and laboratory-based workflows, which can limit its speed and practical application. Recent developments in artificial intelligence (AI) and mobile computing have created new opportunities to automate forensic biometric analysis through portable, real-time digital solutions. This study focused on the development and validation of Foot AI Prediction, an Android-based mobile application designed to perform automated age and gender estimation from digital barefoot prints using artificial intelligence.An experimental research design consisting of two sequential phases was employed. The first phase involved designing and developing the mobile application using a user-centered approach that integrated digital image acquisition, image preprocessing, and an morphopod AI prediction engine into a single Android platform. Digital barefoot print images were captured using high-resolution flatbed scanners, after which the plantar region was interactively cropped before AI-based analysis. The second phase assessed the application's predictive performance by comparing AI-generated age and gender classifications with verified ground-truth demographic information linked to anonymized digital barefoot print samples. Validation demonstrated that the application successfully executed the complete analytical workflow, encompassing image acquisition, preprocessing, AI inference, and demographic prediction, without operational failures. The system consistently processed barefoot print images acquired through flatbed scanners, producing real-time predictions of gender and predefined age groups within a few seconds. Representative validation cases further confirmed its ability to differentiate morphological variations among demographic categories, while the integrated preprocessing module effectively removed background artifacts and standardized footprint images to ensure reliable AI-based inference. The findings demonstrate that integrating artificial intelligence with mobile computing provides a practical and effective approach for automated forensic demographic profiling from barefoot prints. By enabling rapid, portable, and objective estimation of age and gender directly on Android devices, the Foot AI Prediction application offers considerable potential for supporting crime scene investigations, forensic anthropology, disaster victim identification, and other medico-legal applications. Although additional largescale validation using more diverse datasets and authentic forensic casework is warranted, this study establishes a strong technological foundation for AI-assisted barefoot print analysis and contributes to the continued advancement of digital forensic science.