Based On The Design Of The Arm Digital Watermark Trademark Detector

By analyzing the situations home and abroad, aiming at the demand of market and the practical applications, the instrumentation for trademark detecting in digital watermark based on ARM9 is introduced in this thesis.In order to choose robust extraction algorithm of digital watermark, the algorithm of discrete cosine transformation, wavelet transformation and chief feature direction vector which is based on DCT and DWT is simulated with Maltab, including the situation of non-noise and non-attack, some noise, filtering and the RST attacking (translation, revolving, scaling) and so on. By the contrast of the results of simulation, the chief feature direction vector is chosen as the detecting algorithm of digital watermark.In the software and hardware's implement of trademark instrumentation, firstly the hardware of trademark instrumentation is introduced, including the ARM9 processing chip and the periphery expander, the equipment of collection, man-machine connection and so on. Then the lead of Bootload as well as the transplantation of the Linux operating system based on ARM9 is studied. And also, the embedded cross compiling environment is established in order to provide a good platform for design and the research of detecting instrumentation. In the software design aspect, the algorithm of the chief feature direction vector which includes wavelet transformation, the discrete cosine transformation, the Zigzag arrangement and the correlation detection procedure is realized with C language. Moreover, the program of image collection, man-machine interaction and the terminal display of the trademark instrumentation of digital watermark is also elaborated in the thesis.