Adaptive reconstruction of the heterogeneous scan line ETM+ correction technique


ETM+ is a land-imaging sensor with great and wide use in many fields, however, after May 2003, becauseof a technical defect in the sensor´s system scan line corrector SLC, it started giving images of earthcontaining black gap lines at a 22% rate. These gaps made the process of analyzing and extractingaccurate information from these images difficult and complicated. Therefore, scientists have developedmany techniques to remove the gap lines from all ETM+ band-images and complete the missing data. Inthis study, three different ETM+ time images with a 16-day interval between them were used to fill gaplines, complementing the images, and minimize the effect of the gaps. However, the adopted temporalfiltration method was not able to completely remove the gaps. An added algorithm is used, which is basedon using one of three filters (mean, median, and max.) to work on a vertical linear window whose center ison the points of the gap lines in the images. The quality of the resulting images was evaluated using theJaccard metric standard to find the degree of similarity between the original images and the images thatresulted from the implementation of the algorithm, the results showed good quality images, and the gapswere completely removed from the resulting images. A higher degree of similarity was obtained fromimplementing the (mean and median), ranging between 0.6568 and 0.9007, compared with the (max),where the similarity ranged between 0.6246 and 0.8820. This study aims to fill the images gaps using thetemporal filtering technique.