Color Recognition And Its Application On The Intelligent Robot Control System

In the design of robot, a wide range of knowledge is involved, including sensor detection, artificial intelligence, mechanical design and control theory. With the rapid development of electronic information processing technology, target detection and identification capability become one of the most important symbol of the intelligent robot. Robot can aware of their surroundings through sensors. The information obtained is used to achieve the position of the goal and to plan the path. Therefore, robot vision technology becomes a key in the design of robot control system.In this article, the plan of the control system is designed according to the needs of the Martial Arts Challenge. The research is mainly divided into the research of color recognition algorithm and the design of control system. The latter includes design of hardware circuit and software programming. The control system needs to realize many functions, such as sensor data acquisition, image processing, audio processing, servo control, etc. The main content is as follows:The design of target recognition algorithm:The mainly algorithm studied in this paper is the target recognition algorithms based on color information. Firstly, this paper studied the classic algorithm and did experiment on PC and ARM development board. This paper proposed a new look-up table algorithm converting the RGB color space to the HSI color space based on the experiment. Then the algorithm automatically divides threshold by the Otsu method, and to calibrate the target in the image by connecting domain finally.Hardware circuit design of the control system:Hardware circuit design of the control system includes the connection of the ARM the minimum system, the power circuit, the servo control circuit, camera interface circuit, audio interface circuit and network interface circuit.The design of underlying software:Including the introduction of setting up development environment on PC and the ARM development board as well as some design which is in connection with the actual needs. The design is mainly about the underlying application program of some modules, such as serial port, network, audio, camera, I/O, AD, etc..After the design of the system, the improved color recognition algorithm is used to design the system application program. The system was tested through the system installation and commissioning. The test showed that with the movement of the robot, the relative position of the hydrangeas in the image changes. So the robot can find the location of the hydrangea and adjust its movement with infrared sensors. It can complete the task of the target color identification and positioning preferably.