Image Processing Based Speed Measurement and License Plate Detection of Vehicles

The majority of fatalities, injuries, and property damage are caused by traffic collisions. The principal cause of death is frequent excessive speeding. Deep learning techniques are now frequently used for the analysis of traffic surveillance footage thanks to the recent significant improvements in the computation power of common computers. The system for detecting vehicle speed was created utilizing video processing methods. Through the application of image and video processing techniques, moving vehicle velocity is estimated utilizing vehicle speed detection. Real-time speed analysis of video is performed without the use of camera calibration. Moving vehicles are segregated by using masking and frame subtraction methods. In this technique a video is recorded using a fixed-focus camera while taking into account a number of factors. Image Enhancement is used to enhance the qualities of each frame of the video captured using the camera. The vehicle has a bounding box applied to it to identify it, and each frame's pixel change indicates the car's motion. The speed is then calculated using a distance formula while taking Pixel Per Meter into account. Pixel Per Meter is a unit of measurement that is used to describe how much potential image detail a camera gives at a specific distance. The proposed model's outcomes are differed with those of other models. The proposed method is efficient and can determine whether a result is correct up to 96.3% of the time. Results obtained demonstrate the system's effectiveness.