Application And Research Of Robot Vision System Based On DSP

It is needed to process the image of image sensor captured in robot vision control system that based on vision and model identify. The system based on PC is not real-time, expensive and can' t work smoothly in bad environment. Aimed to the default, this thesis uses the TMS320C6416 as core CPU to process static image captured from dynamic image.Firstly, the paper expatiated theory, history, status quo and application of DSP. at the same time, introduced DSP chip choice of the image processing field .Introducing TMS320C6416 DSP chip character , real-time system configuration ,image processing software development environment and character based on DSP. Secondly, The paper discuss robot vision basic hardware configuration and function models of hardware which based on hardware model DAM6416P .The paper also introduce software design and development diagram of machine vision system, which applied to C6416 DSP .Thirdly, in this paper ,we expatiated related technology and theory of image capture ,image pre-processing, target object recognition and position in the target image. Lastly, the paper debated reality of image pre-processing , feature extraction, classify recognition and position algorithms in robot vision system based on DSP and introduced design course which realized host computer control DSP application program.In the paper, we presented and designed a system that can classify and recognize circular ,quadrate ,rectangle, triangle objects and the intension image is the process object of the system. human-computer communication mode is adopted to identify the target objects. In the part of software design, one is PC aspect, the other is DSP aspect. In the PC aspect ,Visual C++ 6.0 is the development platform , in virtue of IEKSDK function library of DAM6416P,through PCI interface function and the library, built a dialog box, through control button on the box command DSP executive correspond to instruct, realized PC control DSP .In the DSP aspect , firstly, extraction intension weight of capture image pre-processing, . feature extraction, then classification identify, in the end extraction center coordinate position ,width ,area and direction etc. info of target object.