Run-time Image and Video Resizing Using CUDA-enabled GPUs

A recently proposed approach, called seam carving, has been widely used for content-aware resizing of images and videos with little to no perceptible distortion. Unfortunately, for high-resolution videos and large images it is not computationally feasible to do the resizing in real-time using small-scale CPU systems. In this paper, we exploit highly parallel computational capabilities of CUDA-enabled Graphics Processing Units (GPUs) in a heterogeneous computer system for accelerating the content-aware resizing of videos and images. The performance results show that our implementation of the seam carving algorithm achieves up to 235x and 30x speed-ups on the computationally-intensive part of the algorithm compared to the single-threaded and the multithreaded CPU implementations, respectively, on the systems tested. The overall resizing operation is up to 7x and 4x faster than the single-threaded and multithreaded CPU implementations, respectively, which demonstrates the potential to resize videos and large images in real-time.