Satellite Image Gap Filling Technique

Abstract

In the current research work a useful gap filling method was presented. Images over
Al Mosel city in Iraq were used as tested images. The linear algorithm for high
resolution (H.R.) image interpolation were based. The difference between the
proposed algorithm and conventional nearest neighbor algorithm is that the concept
applied, to estimate the missing pixel value, is guided by the nearest value rather
than the distance. In other words, the proposed concept selects one pixel, among
four directly surrounding the empty location, whose value is almost equal to the
value generated by the conventional bilinear interpolation algorithm. The
experimental results show that the proposed method demonstrated higher
performances in terms of H.R. and thus is applicable to gray level satellites images.