Image processing is the task of analyzing and manipulating digital image files to create new versions of the images or to extract important data from them. These digital images are represented by tables of pixels, which are RGB values, or in essence, tuples of numbers. Therefore, digital images are simply multidimensional matrices of numbers, which results in the fact that image processing tasks typically come down to heavy number-crunching.

Since images can be analyzed and processed independently from each other in an image processing application, concurrent and parallel programming – specifically multiprocessing – provides a way to achieve significant improvements in execution time for the application. Additionally, there are a number of good practices to follow while implementing your own concurrent image processing program.

So far in this book, we have covered the main two forms of concurrent programming: multithreading and multiprocessing. In the next chapter, we will be moving on to the topic of asynchronous I/O, which is also one of the key elements of concurrency and parallelism.