The phase unwrapping is fundamental in interferometric data analysis but remains challenging in the presence of discontinuities and complex structures. In this paper, what we believe to be a new, fast, and robust phase unwrapping technique is introduced. The method segments the wrapped phase based on gradient polarity, ensuring that each region contains phase gradients in only one direction. By eliminating reversing or conflicting gradients within each segment, the approach reduces complexity, making unwrapping more efficient and less error-prone. Each segmented region is then independently unwrapped, simplifying the process and minimizing error propagation. After unwrapping, the segments are merged to reconstruct a globally consistent phase map while preserving true phase discontinuities from the original data. The method enhances computational efficiency while maintaining high accuracy. The proposed approach was evaluated using both simulated and experimental datasets involving challenging phase structures, such as spiral shear phase maps, and was found to consistently outperform conventional algorithms. This framework offers a scalable solution for applications in optical metrology, SAR imaging, and medical diagnostics.