Feature Extraction of Human Silhouette

Abstract

An approach for discovering significant data features or attributes is feature extraction. Feature extraction is demonstrated by extracting feature points from a silhouette curve. In this study, we present a systematic approach for comparing the silhouettes of human bodies from front and side views of images and real-time pictures. The first step in our silhouette detection approach is picture capture. After acquiring the image, a silhouette and contour detection approach is applied to the image to obtain feature extraction, and the human body contour (profile) of binary images (pictures) may be represented using an efficient (structured) detection technique that is an optimum edge detector known as canny edge detection. The recommended method's feature points were validated by examining changes of the silhouette curve in order to evaluate the effectiveness of the autonomous body feature extraction system. Furthermore, the recovered feature points can be used to compute body dimensions in the future (proportions). Canny's study aimed to enhance the image's edge detectors. As a result, the newly developed system can automate body feature extraction and acquire anthropometric data for a range of applications.