FORENSIC RADIOLOGY: AN INTELLIGENT METHOD OF AGE AND GENDER ESTIMATION FROM X-RAY SCANNED BONES IMAGES USING CONVOLUTIONAL NEURAL NETWORKS

Abstract

Determination of identity is always the most demanding aspect of a person, when only bones are left. Identity can be lost due to natural disasters and incidents. Police departments and other law enforcement agencies engage the forensic department to estimate the age and gender when such an event occurs. In this study, a comparative identification method is being utilized on panoramic dental X-rays through the segmentation of morphological features. Which will be extracted, and our model will be trained using DCN by comparing the atlas and cranial measurements for age and gender estimation, respectively. The proposed model is trained and produced a 98% accurate result during the test phase with 840 images of different age and gender groups. This study is being introduced for the very first time in Pakistan. It will help police, forensic, and other law enforcement agencies to make faster, possible their investigation process in economical. It will help to reduce the total amount of DNA.