The development of a tracking algorithm for ambulance detection using squaring of RGB and HSV color processing techniques
One of the factors that often leads to traffic congestion in cities is traffic lights at a road intersection. Customarily, traffic lights are pre-programmed with fixed timers and does not consider vehicle intensity at intersections. It does not have an intelligent sensor to self-learn the road condi...
| Summary: | One of the factors that often leads to traffic congestion in cities is traffic lights at a road intersection. Customarily, traffic lights are pre-programmed with fixed timers and does not consider vehicle intensity at intersections. It does not have an intelligent sensor to self-learn the road conditions and number of vehicles so that the controller will react based on the traffic information at a particular intersection. Furthermore, emergency vehicles such as ambulance, fire engines and police cars, also face similar problems whenever they reach a traffic light. It is difficult to those emergency vehicles to bypass the congested traffic at traffic lights due to unintelligent traffic light system. Moreover, present smart traffic light system requires to adapt with new algorithms and technologies, for instance vision sensors that are able to detect emergency vehicles that needs to pass by a traffic light. In the future, it is expected that the detection from a tracking algorithm will automatically switch the traffic light signals based on the road conditions. In this study, a tracking algorithm is developed by means of image processing technique in detecting ambulance. Through the combination of two color space, the tracking algorithm can detect the ambulance with higher percentage of detection and is insensitive to the varying illumination of sun light. HSV color space and RGB color has been used to analyze the light of the emergency vehicle as well as a feature for classification. By the combination of morphological approach to select the region of interest, the tracking algorithm delivers promising results as the tracking algorithm achieved 90% of ambulance detection at road intersections. |
|---|