A novel approach employs the principles of medical image analysis using Wavelet Transform (WT) and Difference Peak Signal-to-Noise Ratio (ΔPSNR). Both techniques are combined as a function of different decomposing le...A novel approach employs the principles of medical image analysis using Wavelet Transform (WT) and Difference Peak Signal-to-Noise Ratio (ΔPSNR). Both techniques are combined as a function of different decomposing levels of wavelets and various image search through and slicing levels, which is implemented under MATLAB environment. In this new approach, the structural change due to damage in the component or the presence of foreign bodies appearing in an image taken for a specific structure is uncovered with its extent determined after applying the search through algorithm. Such alteration of the composite structure, which could be masked by the presence of noise, is accounted for using combined WT and PSNR. Effect of Artifacts and Blurring caused by different wavelet types is investigated before choosing an appropriate wavelet, namely Sym8. This new approach, which also reduces the required layers of search within an image, produces a pattern matrix per damaged area and is an excellent way in tracing and modeling damage in structures with ability to predict effects of further damage on components and further application to artificial limbs that could suffer damage and affect users mobility.展开更多
文摘A novel approach employs the principles of medical image analysis using Wavelet Transform (WT) and Difference Peak Signal-to-Noise Ratio (ΔPSNR). Both techniques are combined as a function of different decomposing levels of wavelets and various image search through and slicing levels, which is implemented under MATLAB environment. In this new approach, the structural change due to damage in the component or the presence of foreign bodies appearing in an image taken for a specific structure is uncovered with its extent determined after applying the search through algorithm. Such alteration of the composite structure, which could be masked by the presence of noise, is accounted for using combined WT and PSNR. Effect of Artifacts and Blurring caused by different wavelet types is investigated before choosing an appropriate wavelet, namely Sym8. This new approach, which also reduces the required layers of search within an image, produces a pattern matrix per damaged area and is an excellent way in tracing and modeling damage in structures with ability to predict effects of further damage on components and further application to artificial limbs that could suffer damage and affect users mobility.
