This paper presents an approach for the fault-tolerant computation of the r
ank order filtering on a SIMD (Single Instruction Multiple Data) mesh proce
ssor. The proposed approach improves over a previous approach in two respec
ts: by changing the data dependency in the execution of the rank order filt
ering, a new algorithm with constant execution time complexity can be desig
ned; and by introducing a dependency for the rank values of faulty PEs as c
omputed by neighboring (fault free) processing elements (PEs), a lower dist
ortion can be achieved for the image enhancement
An algorithm is presented; this algorithm is based on a two-phase technique
in which the rank corresponding to each faulty PE is computed by a selecte
d fault free PE. The effect of the rank computation for a faulty PE is rest
ricted within a window of the image, such that no significant overhead is a
ccounted in execution complexity. This results in a considerable improvemen
t in speed-up ratio, fault-tolerance as well as a lower distortion in the r
econstructed image. The proposed approach is also evaluated using experimen
tal results. As the proposed algorithm is directly compatible with a SIMD m
esh execution, a discussion of the communication constructs and switching a
rchitecture to support array processing is also presented.