The state of charge (SoC) is one of the most important and commonly used battery performance indicators in a battery management system (BMS). It helps prevent overcharging and deep discharging, leading to rapid degradation of battery life and deterioration of battery performance. SoC indicates how much charge remains in the battery and is typically used in applications such as electric vehicles (EVs), photovoltaic systems with battery storage, and portable electronics to monitor and manage battery usage. Five general classifications can be used to group battery state of charge estimation techniques: Conventional methods, Observer methods, Adaptive Filter methods, Data-driven methods, and hybrid methods. This paper will present, on the one hand, the different techniques used in lithium-ion battery modeling, and on the other hand, an overview of various SoC estimation techniques, highlighting the benefits and challenges associated with each method. Based on the comparison of studies available in the literature, data-driven and hybrid methods such as the least squares support vector machine (LSSVM) and the extended Kalman filter based data-driven (EKF-XGBoost) enable accurate SoC estimation with lower errors.