During January 2025, contributed to the espressif/opencv repository by developing a RISC-V vector-accelerated phase calculation feature. This work involved implementing a hardware abstraction layer for the cv::phase function, utilizing vectorized arctangent routines for both 32-bit and 64-bit floating-point data. By leveraging C++ and expertise in computer vision, embedded systems, and performance optimization, the contribution accelerated phase computations on RISC-V targets and improved performance-per-watt for embedded vision workloads. The implementation addressed minor integration issues during rollout and established a foundation for broader SIMD optimizations, setting the stage for future enhancements within the OpenCV pipeline on embedded platforms.