Computer Aided Diagnostic of Dry Eye: Toward an Automatic Diagnosis from Noisy Tear Ferning Images
This work focuses on the use of digital image analysis for automatic diagnose of Dry Eye. Recently, scaling of tear ferning images been used to diagnose dry eye. A classification approach in wavelet space is proposed as an initial step toward an automatic diagnose of dry eye. Correlation coefficient, in wavelet space, between a reference image and unclassified noisy tear ferning image, at different level of wavelet decomposition is used as classifier. The noise in the images has a great impact on the value of the correlation coefficient (CF). The noise reduction using wavelet technics at different level of wavelet decomposition (WD) provides a strong improvement for the classification (diagnose) of tear ferning noisy images. A set of reference image representing the scaling of the tear ferning images is used. The CFs between the reference images against each other was used in order to estimate the value of the background noise. The CF was first performed between the reference images and the noisy ones without wavelet decomposition. Then, CFs was obtained for five levels of WD. A comparison of the CFs at different wavelet levels is performed. It shows that CF values highly increase upon increasing the level of WD. This result provides strong evidence that the proposed new approach for automatic dry eye diagnosis is highly promising with the advantage of great accuracy and time processing reduction. Cloud computing is an emerging hypothesis of computing that replaces computing as a personal commodity by computing as a public utility. As such, it offers many advantages in terms of public utility system, in terms of economy of scale, flexibility and convenience. Here many issues rises like not least of which are: loss of control and security. The aim of this study, to understand the different approaches of virtualization and appropriate approach of virtualization in cloud computing.