<p>In this study, we propose a lithium-ion battery state of health (SOH) estimation method based on capacity increment analysis and data-driven approaches. In the first step, actual vehicle battery operational data from a remote data platform are used to calculate a meaningful battery capacity. Subsequently, the capacity increment curve associated with a constant current charging of the lithium batteries is analyzed. A sequence of capacity increments, processed with respect to monotonicity and interpolation, is employed to extract 22 battery aging characteristics, such as mileage and the average value of the capacity increment corresponding to the voltage of a representative battery. The Pearson correlation analysis method is then used to calculate and establish a correlation between these characteristics and battery capacity. Finally, a data-driven model, using the aging characteristics as input, is established to estimate battery capacity, and arrive at a reliable SOH estimation. The effectiveness of this estimation method is validated by comparing the experimental results with the actual battery capacity, resulting in an average absolute estimation error of less than 3.5%.</p>