TRACKING HUMAN EXISTENCE USING CAMERA IMPLEMENTING HUMAN BODY FEATURE EXTRACTION

• Main Author: CHRISTANTO KHUANGGA - NIM : 13514084 , MARTINO
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/28781
• Institution: Institut Teknologi Bandung

<p align="justify">Human existence is an important aspect to assess one's discipline within an organization. However, human error often become the main cause of undetected human existence. This could happen because today detection systems often need special action from the person. <br /> <br /> <br /> In this final project, author interested to implement human existence tracking system using camera, to remove the special action needed. Author propose human body feature extraction as a method of human existence tracking to keep the system remain no need for special action. Additionally, features on human body tend to be easier to detect and serve as a strong identity to mark person who enter a room. <br /> <br /> <br /> This application has two main process. First, detect person enter a room and detect (the same) person leave the room. This application was implemented by using image processing techniques such as HOG descriptor, HSV color conversion, and template matching. Additionally, this system also uses initial processing method such as grayscale and Sobel transformation. <br /> <br /> <br /> The application must be real-time with 300 – 600ms processing time each frame. The application must be able to track every person entering the room and save their leaving time. Tracking failure from this application could happen because the human detection process itself is still could not handle some cases. <p align="justify"> <br />