PURWARUPA SISTEM DETEKSI OBJEK WAKTU-NYATA BERBASIS LAYANAN PESAN SINGKAT

Abstract

The powerful algorithm to detect object movement in development of room monitoring system is very urgent. The commond algorithm needs complex computation. In this research, the prototype of real-time object detection system using simple algorithm is developed, i.e. using the determination of the max noise/pixel value and the tolerance threshold of image accurately, and then the system automatically send a SMS (short message services) to user when the object movement is detected. The developed prototype used a Logitech QuickCam webcam, a Siemens C45 mobile phone and a data cable, and the Borland Delphi 7 with additional components and Serial PortNG Tvideo as system software. The application also includes a database to store the captured images whenever object movement is detected. The test results by varying conditions of light intensities using a 5-watt light bulb, fluorescent lamp 20 and 40 watts indicate that the application is able to automatically detect the presence of moving objects with 100% success rate. The success rate is strongly influenced by the determination of the max noise/pixel value and the tolerance threshold during system configuration. This application is also capable of sending SMS automatically when the system detects a moving object with an average time of 8.35 seconds.