White light interferometry has been applied to micro and nano structure 3D surface topography measurement due to its non-contact, non-damage, large range, high precision and high sensitivity. However, the measurement range is limited by the microscopic objective field. A 3D surface topography splicing method is proposed which can effectively improve the measurement range. Through the domain division of small range and high resolution white light interference pattern collected for many times, the three-dimensional topography data registration method is adopted, and the template matching algorithm is used to match the binary matrix of the image, and the surface topography of the components to be measured is effectively characterized. It is found that when different image splicing methods are used, namely up-down splicing, left-right splicing and 2&2 splicing methods, the effect of image splicing is different, and the accuracy and error are different. Among the three splicing methods, the 2&2 splicing method can effectively reduce the operation error, achieve a large range and high resolution interference pattern splicing, and has a broad application prospect in the white light interference three-dimensional topography characterization.