Can anyone show how to calculate the RMSE (mean squared error) between two rasters images for Landsat TM 5 and Oli 8?
I need to have an efficient method for estimating RMSE for a series of Landsat image ? can we have really an average RMSE?
Dear friends Sattyam Tripathi and Diego Soza,
I thank you for your explanation.
When I viewed the MTL file for Landsat TM5 and OLI 8 images. I noticed that the RMSE is changing. For TM5 the RMSE values are: 0.253; 0.188; 0.189 less than 0.5 other than RMSE values for Oli 8: 3.887; 4.302; 4.24.
For this I ask is what these big differences will not influence the results obtained during the treatments.
Regards
Jad Tahouri
Dear Jad. The L5 L1T product available and accesible trought USGS EARTH EXPLORER, includes in each scene and "GCP.txt" file, wich details the RMSE considering GCP Residual Report under WRS-2 coordinate grid system. If you need a base to calculate the RMSE, the L5 can perform that function to compare it with L8 OLI (not TIRS) scene. I hope this alternative can be a solution for your requirement. Best wishes.
Dear friends Sattyam Tripathi and Diego Soza,
I thank you for your explanation.
When I viewed the MTL file for Landsat TM5 and OLI 8 images. I noticed that the RMSE is changing. For TM5 the RMSE values are: 0.253; 0.188; 0.189 less than 0.5 other than RMSE values for Oli 8: 3.887; 4.302; 4.24.
For this I ask is what these big differences will not influence the results obtained during the treatments.
Regards
Jad Tahouri
Jad Tahouri That is true!...my experience says that's not recommended to apply any geometrical correction to spectral data (as pan-sharp techniques) with no the assumption of an external bias increasing. Then, the "nominal precision" of an image's mission is just a descriptive of the expected horizontal precision, but is not a definitive estimator. Regards.
Diego Soza I wonder ifI can get your help with regards to calculating RMSE for all Landsat-5 TM and Landsat-8 OlI scenes. Can you give me a practical example ?
- Badges
- Science topic
- Similar topics
- Computer Science
- Images