Developed and integrated a FastCV-based hardware abstraction layer for the Canny edge detection API within the espressif/opencv repository, focusing on accelerating image processing for constrained devices. The work involved defining the API surface, specifying constraints for multi-channel input and decimal threshold values, and supporting kernel size and normalization types. By implementing HAL-facing function declarations and definitions in C and C++, the developer enabled offloading of computationally intensive edge detection tasks to FastCV, reducing CPU usage and increasing throughput for computer vision pipelines. This feature laid the foundation for downstream integration and improved performance in hardware-accelerated image processing workflows.