We present a three-dimensional (3D) isotropic imaging of mouse brain using light-sheet fluorescent microscopy (LSFM) in conjunction with a multi-view imaging computation. Unlike common single view LSFM is used for mouse brain imaging, the brain tissue is 3D imaged under eight views in our study, by a home-built selective plane illumination microscopy (SPIM). An output image containing complete structural information as well as significantly improved resolution (～4 times) are then computed based on these eight views of data, using a bead-guided multi-view registration and deconvolution. With superior imaging quality, the astrocyte and pyramidal neurons together with their subcellular nerve fibers can be clearly visualized and segmented. With further including other computational methods, this study can be potentially scaled up to map the connectome of whole mouse brain with a simple light-sheet microscope.