Multi-focus Red and White Blood Cell Image Fusion Based on Content Adaptive Blur

The imaging results of red and white cells in the microscopic images are inconsistent at different focal distances, which may cause an indistinct and incomplete display of red and white cells in a same image. Therefore, in this study, a multi-focus cell image fusion algorithm based on the content adaptive blur (CAB) is proposed to display the solve the incomplete display problem under different focal distances. The content adaptive blur method is firstly used to detect and segment the focus area in the multi-focus cells images automatically, obtaining the initial segments of image. Then, a refined segmentation of the foreground and background at a higher accuracy is applied by using graph cut to obtain the decision map. Finally, the final fusion image can be obtained by performing the pixel-level fusion technique. Results show that the algorithm proposed can completely and clearly display red and white cells at different focal lengths. The fusion image on image contrast, structural similarity, image clarity and image quality in this study presents a better result with comparison of other state-of-the-art algorithms.