Control System Based On Visual Sensing Cricket

During the process of producing and living, human beings have approached the limits of physical and activity territory that they cannot reach by taking advantages of machines. In company with the development of Image Processing, Image Recognition and correlative subjects like Pattern Recognition, Artificial Intellegence, it is no longer a dream that intellectual machine has the ability of understanding the information of the environment by applying vision like humans, making decisions according to such information and acting quickly. This thesis studies real-time recognition and spatial position acquisition of moving object based on plain background upon the Ball-Plate System, and simulate the process of a machine vision system understanding environment and objects.The thesis establishes the camera model, studies and implements calibration method of linear camera model, and can figure out the actual position of the object in the world coordinate according to the result of calibrating process. Studies the process of recognizing moving object and calculating the actual position of the moving object in single environment, works out a process of image processing, information acquiring and feedback, as well as the method of coordinate transferring of the Ball-Plate System. As long as getting the position of moving object in 3-D space is concerned, how can the Ball-Plate System get the position of moving object, this thesis offers a method of obtaining the actual position of the object in the environment according to information offered by the 2-D image along with the information of the plane on which the object is moving during the control process; implements Real-time recognition of moving object, and both the precision and velocity of such process can fit the requirement of the control system.Combined with the linear model camera calibration method, this subject implements obtaining the real position of the moving object as feedback to the control system, which is applied in the real system, and the position control of the moving object can be realized by this method.