Computer Vision-Based Approach for Estimating Age andGender using Wrist X-Ray Images

A crucial phase in the legal procedure, notably in biometrics that deals with legal and judicial law, is estimating a person's age and gender. The hand, skull, and teeth, which have more distinguishing characteristics and components, can be used to estimate age. X-ray images frequently do bone growth assessments. Consequently, it was decided to include x-ray images of the hand in the imaging process. Preprocessing, male and female classification and age prediction are the different sections of the proposed study. The wrist bone picture will undergo Otsu thresholding, and edge-based segmentation is used to identify the region of interest. The wrist bone image is then input into the CNN classifier for gender classification and the Logistic Regression model for age prediction. The proposed method chooses the wrist bone's performance metrics based on age and gender. Most of these measures are taken directly from the bone. An RSNA bone age dataset contains 12,612 left wrist x-ray images. The likelihood that the classification is accurate is 95%. The most crucial elements in this inquiry are the participants' ages and genders. The convolution model based on the feature extraction performs better in the gender classification and age prediction models, thus enhancing overall accuracy for image-based age and gender classification assessment.