Smart V-the Agv Binocular Vision System

Machine Vision is the current research in the field of intelligent robots as a hot spot. With the improvement of automation and intelligence,a lot of the practical systems need to add the functions of machine vision.This article is based on the vision-based automatic guided vehicles(V-AGV)of the binocular stereo-vision technology, two color cameras are installed on the AGV system.The system use the color camera images to identify interesting objects to identify and positon the target,and provide navigation information for the vehicle.First of all,the paper reviews the the development and the research status of the AGV system in the world,Color images are the basic information of V-AGV vision system to sense the surrounding environment. How quickly and accurately to identify the target is the primary problem.Secondly,pre-process the collected color images,and use the HSI color model and geometric shape recognition methods to identify the target from the images,and improve the recognition accuracy.Finally,the camera calibration and three-dimensional information recovery techniques of the V-AGV vision system are the focus points.In camera calibration, we use Three-dimensional calibration template to calibrate the internal parameters and the external parameters of the cameras.This method is based on the pinhole camera model,use the transformation matrix of perspective projection to calculate theparam-eters of the cameras. In the world coordinate system, the three-dimensional calibration template is used as an intermediary to get the relate position between the two cameras. In order to get the three-dimensional coordinates of the objects, we use the least squares method to solve the projection matrix equations and obtain the three-dimensional coordinates of the target objects. Experimental results show that the approach is reliable.