This paper presents a fault detection and classification strategy based on the wavelet transform (WT), utilizing current signals in smart transformer (ST)-based hybrid microgrids. The Haar wavelet is chosen as the mother wavelet function to extract features from current signals due to its effectiveness in capturing critical fault characteristics. Fault detection is performed by identifying the maximum values of the detail coefficients of the AC currents, while fault classification is achieved with the help of the maximum detail coefficients of the three-phase AC currents along with the ground current. The proposed method demonstrates reliable performance under various fault scenarios, including different fault types, fault location, fault resistances, and ground resistances. This approach is effective in detecting both symmetrical and unsymmetrical faults, ensuring accurate classification across diverse fault scenarios. The proposed method is validated on an ST-based LV meshed microgrid model developed in the MATLAB/Simulink environment.