Abstract | ||
---|---|---|
Invalid image pixels often cluster within the overlapping region of neighboring Landsat scenes and can have negative impact on the quality of image composites generated from multiple Landsat scenes. We propose a method to automatically screen invalid image pixels in Landsat-4, 5, 7 images by generating a polygon smaller than the bounding box of each scene. Our assessment results show that masking invalid pixels with the proposed method can effectively eliminate the edge effect and significantly improve the quality of composites. This automatic invalid image pixel screening process can serve as an image preprocessing step to facilitate large area applications with Landsat images based on cloud computing platforms. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1109/IGARSS.2016.7729615 | 2016 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS) |
Keywords | Field | DocType |
image processing, cloud computing, image compositing | Computer vision,Polygon,Feature detection (computer vision),Masking (art),Computer science,Remote sensing,Image processing,Preprocessor,Artificial intelligence,Pixel,Cloud computing,Minimum bounding box | Conference |
ISSN | Citations | PageRank |
2153-6996 | 0 | 0.34 |
References | Authors | |
2 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Qingling Zhang | 1 | 17 | 3.01 |
Kai Chen | 2 | 0 | 0.34 |
Qian Jing | 3 | 0 | 0.34 |
Xi Chen | 4 | 18 | 6.51 |