The observation of three-dimensional (3D) structures of objects provides much more specific information to the biomedical research. As a new kind of 3D fluorescence microscopy, Light-Sheet Fluorescence Microscopy (LSFM) uses a thin light-sheet to selectively excite the fluorescence signal of samples from the side, and collects the signal in the orthogonal direction, enabling its intrinsic optical sectioning ability. Compared to the conventional Epi-fluorescence microscopy, LSFM only illuminates the in-focus area and minimizes the phototoxicity and photobleaching, making it very suitable for long-term imaging for living specimens. Besides, LSFM uses CMOS or CCD sensors to capture the image plane-by-plane, possessing the merit of high-speed imaging rate. Therefore, LSFM has been broadly applied in the fields such as cell biology, neuroscience, developmental biology, etc. However, commercial light-sheet fluorescence microscopes are normally bulky and expensive, and thus are usually placed in the shared equipment center where researchers have to queue up for their time-consuming experiment appointments. The delicate biological samples may degrade or expire during transportation, which is negative to the investigation. What's more, to improve the image resolution of LSFM, many researches have managed to combine LSFM with other super-resolution microscopies, such as Structured Illumination Microscopy (SIM), Single-Molecule Localization Microscopy (SMLM), but at the cost of further increasing the size and the expense of the system.In this paper, we designed and built a compact multicolor light-sheet fluorescence microscope. In the illumination path, a 1D scanning galvanometer is used to scan the Gaussian beam to generate the Gaussian light-sheet, and a Dove lens is added to change the orientation of the light-sheet to make it coincide with the focal plane of the detection objective. In the detection path, a detection objective and a tube lens are used to collect and image the fluorescence signal with the sCMOS camera. The volume of the system is only 30 cm(L)×30 cm(W)×40 cm(H), which is merely about 1/5 of that of the Zeiss's Lightsheet7 light-sheet fluorescence microscope. To improve the spatial resolution of the system, we used the Mean Shift Super Resolution (MSSR) algorithm to process the acquired data. As a computational super-resolution method, MSSR can recover a super-resolution image from a single frame of fluorescence image without the need of additional optical and mechanical components, facilitating the spatial resolution improvement of the compact light-sheet fluorescence microscope. The lateral resolution of the system is 530 nm, which can be improved to 330 nm after using the MSSR algorithm. We conducted imaging experiments on the spinach root slice and the rat intestine slice. With the 3D multiple field-of-view stitching technique, a dual-color 3D fluorescence image of the rat intestine slice is achieved with the field-of-view of 2.75 mm×3.35 mm×1.55 mm (4 180×5 100 pixels@850 slices). The results show that the system can realize super-resolution fast multicolor 3D fluorescence imaging of biological samples, providing a useful microscope for the biomedical researchers.