Spatial resolution is a crucial index to measure Brillouin optical time domain reflectometry. The spatial resolution of traditional Brillouin optical time domain reflectometer is restricted to 1 m, which cannot fully meet the demand for high spatial resolution in the engineering field. To improve the spatial resolution of the system, researchers have generally used two methods: optical pulse modulation at the optical path end and signal processing at the circuit end. However, the method of back-end signal processing has certain advantages in simplifying the system structure, reducing the hardware cost, and realizing the iterative upgrade. Furthermore, it can meet the requirements of the instrument's cost-effective and batch industrial applications in the engineering field. This study introduces comprehensive signal processing methods based on pulse segmentation and Brillouin spectrum analysis to improve spatial resolution. The principle and experimental scheme of these methods are analyzed and discussed, and the results obtained by each method are summarized and compared. Finally, the development prospect of Brillouin optical time domain reflectometry is prospected.