Research On DSP Technology Tungsten Ore Primary System

China is rich in mineral resources, especially in the reserves of tungsten ore resources hasthe great advantage of the unparalleled wolframite reifning the most important ore of tungsten, itis also known as wolframite. Of tungsten ore enetrprises in the primary process of tungsten oregenerally through artiifcial hand-selected to complete the the slag separation of work，it is alsothe most important tungsten ore primaries. Obviously, the primary way of tungsten oreproduction efficiency is low, and the subjectivity of the larger objective unity is not conducive tothe production standards, the technical level is lagging behind. Along with the rapid developmentof the image processing technology and DSP processor performance, the image processingsystem under a variety of versatile DSP platform has been widely used. Thus on the basis ofrational analysis for the wolframite visual characteristics, combined with machine vision theoryand embedded processing platform design based on DSP technology the tungsten ore primariessystem. This paper is as follows:(1)Analysis of the current tungsten ore primary process, the stage of development of thedigital image processing technology, Research, focused on DSP technology development processand the structural features of the DSP chip oriented papers structures of the hardware platformand tungsten ore recognition algorithm designed to provide a theoretical basis.(2)Discussed in detail the part of the entire tungsten ore the primaries system hardwareplatform structures and various functions. Including mainly video decoding circuitry, a videoencoding circuit, data storage circuit.(3)Designed to improve the global threshold segmentation algorithm for image processingas the main mineralization positioning algorithm, the completion of the run-of-mine imageacquisition, image feature extraction, identiifcation and localization of mineralization thetungsten ore primaries system.(4)Discussion DSP code optimization strategies, and target systems in image processingalgorithms to use a lot of code optimization, greatly enhance the processing performance of thesystem.Experimental results show that the system can achieve real-time, non-contact intelligenttungsten ore primaries, small size，high stability, low power consumption, and low cost at thesame time, with a high practical value.