This work evaluates the acceleration of image-processing filters on multicore CPUs using the OpenMP parallel programming model. Sequential and parallel implementations were analyzed across different resolutions and filter types to measure speedup, efficiency, and scalability. Results show that OpenMP significantly reduces execution time, especially for high-resolution images, where parallel processing compensates for thread-management overhead. The study also demonstrates the feasibility of integrating OpenMPbased preprocessing into AI pipelines, confirming that multicore CPUsremain a viable and accessible solution for real-time computer vision tasks.