Faster image filtering via parallel programming

Various computer methods are sourced in parallel programming. Advances in methods and techniques with their appropriate usage are beneficial for multimedia applications. Parallelization can significantly decrease the time of calculations. In this article, we analyze how the speed of calculations is influenced by the usage of parallel algorithms in image filtering processes. Additionally, we define a novel cluster selection method which helps in more efficient processing the images. We present a method based on multithreading and the division of the image for rectangles, while as cluster selector we have used k-means algorithm. Proposed filtering is applied parallel on each part of the image, where each of threads is working for each task. Results show that our proposition can give positive results and faster filtering of images when compared to the classical approach. Furthermore, the algorithm connected with k-means and filtering can be very helpful in objects detection.