The Research On Embedded Vision System Monitoring Oriented To The State Of The Object

With the advancement of the technology and development of needs, the visual monitoring systems are developing towards to the direction of intelligent, networked, high speed, and high accuracy. In the modern industrial, manual or monitoring as the core of IPC have diminished meet the requirements of the monitoring when monitored the various status information of products on the line. In the view of lower efficiency and integration of the equipment, This thesis proposed a visual system combined with embedded system. The system could output the control signals through acquiring, processing the state information on the products. It could achieve the purpose of monitoring the state information of the products.According to consider with the application feature of kinds of embedded systems, the overall design is setted. The system is composed with hardware and software. The hardware store and operate the data when the system is working. The software is responsible for compiling the program building and algorithm achieving. According to analysis and comparison with several DSPs, TMS320DM642 of TI has made certain to be the core chip of the embedded systems.The hardware keep to the principle of minimum system. As DM642 to be the core of the system, the thesis has completed the chip selecting and designing the circuit of video encoder and decoder, external memory, serial output, JTAG and the power supply that based on the requirement of interface and process of the visual system. And the circuit diagram of the interface has been drawn.The software which based on the hardware has been analysised. In the development environment of CCS and the DSP/BIOS operating system, the program of the initialization of every parts has been programming. In the view of the problems during software developing, the solution has been given.This thesis gave two examples, which have universal significance, to achieve monitoring the status information of the products by embedded vision systems. One is based on the colour of the products; the other is based on the shap of the products. The experiments indicate that this embedded vision systems could monitor the status information of the products quickly and accurately.