Video monitors traffic generally does not detect the speed, quantity, and length of passing vehicles. This research provides the design, manufacture, and analysis of vehicle detection system with computer vision technology using OpenCV library. The system can display the output of the number, speed, and length of the vehicle. The study was conducted using OpenCV as a library of C++ programming language. In this research tool used is a web camera mounted on the pedestrian bridge. Web cameras will record the movement of the vehicle at a certain distance. To detect the movement of vehicles, we need a background image. Background image was obtained by using the method of Time Average Background Image (TABI). Reduction of pixels between the background image with the image captured by the web camera will produce a vehicle is detected. The vehicles will be marked with a red box. The speed of a vehicle can be measured by dividing the distance traveled by the vehicle travel time of vehicles. This system has been successfully calculate the speed of a passing car with an error range of about 2.44 to 4 percent. The number of vehicles can also be calculated with an error of 26.7 percent, and the length of the vehicle can also be calculated with an error of 11%.