Among the numerous channel estimation (CE) techniques in the literature, pilot-aided CE has been widely employed, where interpolation methods can be used for reducing pilot overhead. The conventional family of linear interpolation (LI)-based algorithms is easy to implement, but naturally results in residual interpolation errors even at high signal-to-noise ratios. In this letter, an enhanced adaptive polar LI method is proposed, which exploits new algorithms of effective origin optimization, adaptive angle gradient, and sparse channel detection. The proposed scheme is capable of outperforming a range of existing methods and adapting to any tap-delay-line type channel model, in real-time, while maintaining a complexity as low as the traditional LI method.