Based On The Digital Image Processing Of Wood Defect Recognition System On DSP

Now the DSP technology widespread use in image processing and provide better technology to the image processing technology, with DSP as the core of the digital image processing technology will also become the future trend of image processing, so this paper will show lumber image processing results in DSP on implementation.(1) Based on the knowledge of image processing, wood defects knowledge as well as the DSP technology knowledge elicitation based algorithm. This paper uses region growing method and support vector machine to the method of combining the defective portion of image segmentation, the improved method with the region growing segmentation and SVM segmentation image comparison, its effect is more obvious, segmentation part is more accurate(2) Lumber image processing is the core part of the image segmentation, in order to make integral effect and method is more simple, this paper to defect image processing the marginal, Laplace operator application of image marginalized, make defects part be clear at a glance, and make segmentation part clear, easy to contrast.(3) The ultimate goal of wood defect system is to come true the defect characteristics and species identification, this paper,first of all is to find out a suitable classification method, and to show the experiment results. The classification results are analyzed, to reduce the classification error.(4) On the DSP chip performance in-depth understanding, especially for the C6201chip, the chip specific internal structure and basic peripherals and so on, the central processor core components of knowledge presentation, on chip using method and scope of understanding, increase the knowledge about the core part of the hardware.(5)The paper through the DSP system algorithm, first in MATLAB environment of the algorithm is achieved, then the algorithm to DSP system transplantation. Through a system of DSP, introduces the general idea of the whole system composition and working mode and process. In the system of possible problems were introduced, and given some processing method. The image processing system shows more satisfied effect.