Route Planning System Of Smart Vehicles Based On Binocular Vision

In industrial and agricultural fields,people cost much manpower,material and financial resources to carry out real-time monitoring,so the existence of mobile robots will be a good substitute for human.Among them,intelligent vehicles have the characteristic of efficiency,flexibility and mobility,being the best choice for mobile robots.As a substitute of the intelligent vehicles,the intelligent cars are cheaper with simple circuits and easy to control and debug.Besides they can also do obstacle avoidance and path planning.In this paper,we build an intelligent car path planning platform.In the hardware part,the car uses Arduino as a controller,L298 as a motor driver chip and two network cameras as sensors.Besides,the cameras can also do information transmission and receiving control instructions.After changing the network cameras wiring,the car can be controlled by computer.Visual Studio 2013 and Opencv function library are used to extract feature,camera calibration and binocular vision based on computer vision technology,so as to generate environment virtual map.In the path planning aspect,not only the specifications and dynamics constraints of the trolley are taken into account,but also the potential field model function is improved(that is,the relative distance is added in the repulsive field and the threshold is added in the gravitational field)to prevent the car from falling into the local minimum value and failing to reach the target point.Furthermore,based on the distance information provided by the disparity map,a gray value threshold is set to limit the minimum distance between the moving car and the object,to ensure that the car can be safely braked within this distance.When the car is close to the distance,the car will turn to prevent collision with the obstacles.After several Matlab simulation experiments,it can be seen that the proposed algorithm can not only measure the distance more accurately,but also avoid the occurrence of shock and plan a collision-free path quickly.The intelligent car path planning platform built can be used in generating practice and have application and reference values.