Prototype of Bridge Navigational Watch Alarm System Equipped Obstacle Warning System Based on Image Processing and Real-Time Tracking

Abstract

Ships are sea transportation that is often used in Indonesia. However, who would have thought that the transportation would become a case because of the frequent occurrence of accidents. This has been proven from data from the KNKT (National Transportation Safety Committee) which noted that from 2012 to 2017 there had been an increase in accidents in the waters. In fact, according to the National Search and Rescue Agency (Basarnas) in 2020 there have been 878 incidents with victims reaching 4658 people. In this final project, the author makes a prototype of BNWAS (Bridge Navigational Watch Alarm System) equipped with obstacle warning using image processing with the otsu method strengthened by thresholding-based segmentation with inverse technique (TsTN), distance detection using the triangle similarity method, real-time tracking with GPS (Global Positioning System) and the entire system can be observed on the Android application. The Final Project performs several analyzes including performance analysis by calculating the accuracy of the BNWAS alarm system, image detection accuracy, distance detection accuracy, GPS accuracy, overall system testing accuracy and packet loss. The accuracy of each system is very good because the error is below 2%, while for overall system testing has a very good performance with a delay of 179.8 ms and 0% packet loss.