Advanced Digital Signal Processing (DSP) technology provides a solid foundation for the machine vision development, applying it to design a small and intelligent real-time monitoring system about the steel plate can improve the steel production and the economics efficiently. In this dissertation, the design of online steel identification and tracking system about ADSP-BF561, part of the hardware modules, software frameworks, the algorithms about image processing and image transmission are studied.This paper analyzes the importance of machine vision technology use in the steel industry; combining with DSP and monitoring technology, the online steel identification and tracking system is designed; algorithms about steel plate image preprocessing, detection and tracking are studied, focusing on the improved segmentation algorithm for symmetric difference and Kalman tracking algorithm; related theories about image compression and network transmission are studied; at the end of this paper, the client inerface is designed and the correctness of programming on the ADSP-547 platform, the experimental results and the various commissioning results are analyzed and summarized. |