A planar flow cytometric microimaging system is mainly used for cell recognition and classification in urinary sediment and gynecological secretion analysis. The depth of field (DOF) of the microscope seriously restricts its imaging range in the direction of the optical axis, rendering it incapable of imaging all the cells in the whole laminar thickness of the planar flow cytometric microimaging system. In this paper, the DOF is extended by using dual sensors with a common light path, and imaging of high-speed moving cells at a large DOF is realized, thus solving the difficulty that the multifocus super-depth technique can only be used in a static observation sample. A fusion algorithm based on saliency detection and multiscale image decomposition is developed to fuse the dual-depth-of-field images. The multiscale image decomposition uses L0 smoothing for multiscale image decomposition. L0 smoothing is particularly effective in sharpening major edges by increasing the steepness of transition, while eliminating a manageable degree of low-amplitude structures. It can globally control the number of non-zero gradients that result in an approximately prominent structure in a sparsity-control approach; this does not depend on the local features, instead it locates important edges globally. Experimental results show that our approach can enlarge the DOF 1.89 times, and the dual-DOF fusion algorithm can fuse two images with different DOFs into one image with clear multiple targets.