Detection of Vehicle Speeding Violation using Video Processing Techniques
B. Krishnakumar, K. Kousalya, R.S. Mohana, E.K. Vellingiriraj, K. Maniprasanth, E. Krishnakumar
Abstract
Due to their excellent development skills and low cost, detection of car speeding violations using video processing techniques is known to be one of the better applications for Traffic Speed Control (TSC). These systems apply video processing techniques to camera outputs to obtain the required information. This work describes a method for measuring speed of the vehicle depending on motion data. This paper offers a new motion detection-based vehicle speed measurement method. Unlike the methods that are based on features which rely on the characteristics of vehicles such as license plates or windshields, the proposed methodology can find a vehicle's speed by evaluating its operating features for a specific areas of interest (ROI) with unique dimensions. This proficiency allows for real-time computation and outperforms feature-related methods. Vehicle identification, tracking, and measuring of speed are the three main modules in this system. The compound-off-Gaussian background subtraction procedure detects every moving object when it enter into its ROI. Then using morphological changes, the unique parts of these materials become integrated filled forms and some limited filtration functions produces the most likely objects to be a vehicle. The identified vehicles are then monitored by using the bulb tracking method and their displacement in successive frames is determined which will be used in the final speed calculation module. A photo of the car, its associated speed, and the detection time are all included in system releases. In comparison to existing speed measurement techniques, experimental findings show that the proposed technique has reasonable precision.