Unsharp masking operations of original video signals are performed by a couple of processor systems (100, 200), which are interconnected in cascades logically to share the burden of the operations. One processor system (100) in the preceding stage executes a part of unsharp masking operations by accumulating the original video signals with respect to the pixels involved in an original image to transfer the resultant data together with the associated original video signals to the other processor system (200) in the subsequent stage. The other processor system (200) receives those data and proceeds the remaining operations required for the unsharp masking operations including averaging by simple or weighted addition. Those operations are executed with respect to all unsharp masks involved in the original image. Since the one processor system (100) shares the unsharp masking operations with the other processor system (200), the total operational speed is increased approximately twice as much as an image processing system composed of a single-processor system, giving rise to a real-time processing of unsharp masking of images.