In this study, we developed methods for accurate and rapid cancer screening, built a whole slide scanner for appropriate cancer diagnosis, and analyzed cancer growth via image processing. In the first experiment, we used our program to automatically calculate the three-dimensional nucleus-to-cytoplasm (N/C) ratio, thereby reducing the ambiguity during pathological diagnosis. In the second experiment, we performed autofluorescence image, on normal tissue, tumor, and dysplasia specimens. In the third experiment, we performed whole-slide imaging on histological sections of oral cancer slides using automatic scanning system.