Traffic Light Automation with Camera Tracker and Microphone to Recognize Ambulance Using the HAAR Cascade Classifier Method

Abstract

Lack of knowledge by road users regarding these priorities, especially when there is a passing ambulance that is often stuck in traffic at a crossroads due to accumulated vehicles and the traffic light is still red. The purpose of this paper is to simulate traffic light automation by giving a green light every time an ambulance passes by using the HAAR and Computer Vision methods. The HAAR method is used for training data from less sharp images as part of the Ambulance object classification process. The Computer Vision method is used as a tool in image processing objects to processing the image captured by the Camera. Hardware through the microphone performs pattern recognition to pick up ambulance sirens. The test result at the average frequency caught by the microphone is 1.3 kHz. The test results of the System to capture ambulance objects received a precision value of 75%, a recall of 100%, and an accuracy of 75%.