|
A method for normalization and co-registration of multi temporal imagery for change detection
|
Yousef Rezaei *   , Mohammad Javad Valadan Zouje , Mahmood Reza Sahebi |
| Bu-Ali Sina University |
|
|
Abstract: (3162 Views) |
| The multi-temporal Remote sensing data are unique tools for monitoring and detecting land cover change over time. Radiometric and geometric consistency among these multi-temporal data are difficult to maintain, due to variations in sensor characteristics and view, solar angle, and atmospheric conditions, and these variations can obscure surface change detection. The radiometric normalization and geometric co-registration of multi-temporal satellite imagery of the same terrain is often necessary for land cover change detection, e.g., relative differences. In previous studies, in order to obtain radiometric correction of multi temporal imagery, the ground reference data or pseudo-invariant features (PIFs) were used. Using the ground reference data collection is costly and difficult to acquire for most satellite remotely sensed images and the selection of PIFs is generally subjective and need the user’s supervision. In this research, we demonstrate a method for radiometric normalization and geometric co-registration between multi temporal images of the Alam-chal Glacier. The selection of PIFs has been done statistically, and the satellite images are normalized radiometrically to a common scale. In order to image co-registration, first the noise was removed and then repeatedly two images were registered using polynomials models and image matching. The proposed method was evaluated by histogram comparison, statistical parameters and independent check points. The results show that the statistical parameters of two image are nearly the same and the total RMSE of check points was 0.52 pixel. |
|
| Keywords: Multi temporal imagery, radiometric normalization, co-registration, Pseudo Invariant Feature |
|
|
Full-Text [PDF 1385 kb]
(1259 Downloads)
|
Type of Study: Research |
Subject:
Aerial Photogrammetry
Received: 2016/04/30 | Accepted: 2016/10/9 | Published: 2018/06/21
|
|
|
|
|
|
