The Application Of Embedded System Among Robot Vision

An embedded system is a special computer system with variable software and hardware. It is based on computer technology and mainly served for application. It is applied to the system which has strict requirements in functions, reliability, cost, volume, and power. Robot vision system is an important application of embedded system in robot study field. As the function of human's vision system, the robot vision system endows the robot an advanced sensor, which can make the robot respond to surrounded environment in "intelligent"and flexible manner. The purpose of robot vision system is to feed back the status and position information of target or itself by the way of imagine location and imagine apprehension. The purpose of the paper is to develop an all-purpose embedded system platform, on which the operation system can be studied and ported. With the application of robot vision navigation, intelligent robot tracing still target in simple environment can be realized. Based on this system, the embedded system, robot vision and navigation, and intelligent control of robot can be studied. Furthermore, it is a well-developed platform for the purpose of update and development of the system. Firstly, the paper analyzed the architecture of embedded system, and discussed the key technology on its design and development course. Secondly, for the specific application of robot vision, we made the requirement analysis of embedded system and divided the system into different function modules using the method of software/hardware co-design. We selected the S3C44B0X embedded microprocessor and configured image acquisition and motor driver peripheral devices to set up the hardware platform of robot vision system. Also, we made an in-depth study of embedded operation system uC/OS-II which we have used and analyzed the core architecture and implemented the porting for the specific hardware platform-S3C44B0X embedded microprocessor. Finally, based on robot vision software/hardware platform, we programmed imagine collection driver and image matching algorithm complied with the requirements of imagine processing in embedded system. Because we have used modules method in course of system design and implement, the system has opening structure and easy to update and extend. The result of experiments and tests shows that the robot vision system can perform properly and reliably. It can trace still target in simple environment, and achieved the anticipated aim of us basically. Furthermore, the result of research can stretch to vision monitor, intelligent traffic system, and intelligent toy etc. fields and have widely applying and flourishing prospect.