| Summary: | 碩士 === 國立臺灣科技大學 === 醫學工程研究所 === 101 === Histopathology is the study of the microscopic anatomy of cells and tissues using microscopy techniques or electron microscope. During the past many decades, the pathologist sliced the cells or tissue into very thin layers that are mounted on a glass slide and examined under a microscope. With the recent advent of digital imaging, high-resolution digital images are necessary for accurate diagnosis, and it is expected to play a revolutionary role in future histopathology. In the past, biomedical image database can be only stored string data such as DNA sequences and low-resolution images. Most existing large-scale image techniques have bandwidth and storage limitations, so images must be lossy compressed to solve this problems before transmission and storage [1]. It is difficult to display high resolution microscopic images over internet in real time. This thesis presents two ultra fast super high resolution image cloud systems, which enables users to interactively view large, high resolution images within existing web browsers in real time. In evaluation, the method1 is demonstrated to render large high resolution image over internet in real time (0.5~0.67 second on average), and the results show that the method1 performs 3~4 times faster than method2 in rendering large high resolution images while costs approximately half of storage space method1 requires. However, the encoding time of method2 are 1.4 times faster than method1 due to following reasons, the regular encoding of the method2, and the complicated structure of the method1, but it is unnecessary to encode image dimension to 1*1 of the method2.
|
|---|
