Three-dimensional entropy vector median filter for color video filtering

R. Lukac,B. Smolka,K. Plataniotis,A. Venetsanopoulos

DOI: https://doi.org/10.1117/12.503178

2003-06-16

Abstract:We provide a new non-motion compensated adaptive multichannel filter for the detection and removal of impulsive noise, bit errors and outliers in color video or color image sequences. The proposed nonlinear filter takes the advantages of the concept of the local entropy contrast and the robust order-statistics theory. The new entropy based vector median is computationally attractive, robust for a wide range of the impulsive noise corruption and significantly improves the signal-detail preservation capability of standard vector median filter. Because the precision of statistical operators such as mean or entropy increases with the increased number of observed samples, the used spatiotemporal cube filter window guarantees a high accuracy of the proposed method that is able to achieve excellent results in terms of commonly used objective measures and clearly outperforms standard vector filtering schemes