Research On Target Recognition And Operation Technology Of Autonomous Mobile Robot Based On Vision

With the continuous improvement of the level of scientific research and the people's living standard, the research of mobile robot based on the machine vision is more and more popular. The data collected by the image sensor plays a huge role in effectively understanding the environments and objects around the mobile robot.what's more,it will be helpful for target recognition and operation in order to complete the tasks specified by the user. The object recognition and manipulation techniques of autonomous mobile robot under the indoor environment were studied.With combining VS2010 and opencv computer vision library to design host computer software, the intelligent control system based on the quadruped robot platform is capable of multi-sensor data acquisition, wireless communication,remote control of the robot and manipulator, human-machine interaction based on voice, image processing, target recognition and the target center extraction. The main work done in this research is as follows:The machine vision and image processing technology are discussed. With comparing the RGB and HSV color model, HSV model which is less sensitive to the strength of the light is chosen to conduct the color image segmentation.With image preprocessing, object recognition and the smallest outer rectangle, the centroid point can be found.On analysis of the functions and tasks of the system, the software and hardware of the robot control system are designed. The robot is equipped with infrared distance sensor, vision sensor, infrared body sensor, temperature and humidity sensor, gas detector, collision detection sensor. The implementation method of host machine and slave machine software is given. Under a wireless LAN environment,the wireless communication of the client/server mode is realized by the use of socket interface technology based on TCP/IP protocol. Wireless communication platform is built to realize wireless transmission of instructions and to collect environmental information,to acquire and process image as well as to recognize object in order to control the robot remotely.Human-machine interaction system is designed and the speech recognition and speech synthesis technology and realization method are respectively introduced. Combining speech technology and robot control technology, the robot remote control is more convenient and intelligent. In addition, speech man-machine interaction and remote control experiments based on platform are designed to verify the reliability and easy operation quality of the phonetic system.Combining infrared distance sensors to detect the range between the robot and the panel, the strategy to approach target is made for autonomous robot based on the target recognition and extraction of image centroid point. The robot stops after arriving at the position where the end of the manipulator can reach the object. Furthermore, the kinematics equation of the three degree of freedom manipulator is illustrated to obtain required angle for driving the Steering gear, so as to control the actuator to arrive at the desired position to manipulate the target.