Connectionist median filtering networks

摘要

A family of connectionist rank filtering networks is presented, suitable for real-time median filtering. It is demonstrated that these have low complexity (from O(N2) processing elements and connections for a single output N element filter to O(N2) for a fully parallelised N × N filter). The networks have some unusual requirements in terms of processing elements and topology compared with the standard neural network models. The feasibility of VLSI implementation of the networks is discussed.