Integrated Preprocessing of Multitemporal Very-High-Resolution Satellite Images via Conjugate Points-Based Pseudo-Invariant Feature Extraction

Abstract

Multitemporal very-high-resolution (VHR) satellite images are used as core data in the field of remote sensing because they express the topography and features of the region of interest in detail. However, geometric misalignment and radiometric dissimilarity occur when acquiring multitemporal VHR satellite images owing to external environmental factors, and these errors cause various inaccuracies, thereby hindering the effective use of multitemporal VHR satellite images. Such errors can be minimized by applying preprocessing methods such as image registration and relative radiometric normalization (RRN). However, as the data used in image registration and RRN differ, data consistency and computational efficiency are impaired, particularly when processing large amounts of data, such as a large volume of multitemporal VHR satellite images. To resolve these issues, we proposed an integrated preprocessing method by extracting pseudo-invariant features (PIFs), used for RRN, based on the conjugate points (CPs) extracted for image registration. To this end, the image registration was performed using CPs extracted using the speeded-up robust feature algorithm. Then, PIFs were extracted based on the CPs by removing vegetation areas followed by application of the region growing algorithm. Experiments were conducted on two sites constructed under different acquisition conditions to confirm the robustness of the proposed method. Various analyses based on visual and quantitative evaluation of the experimental results were performed from geometric and radiometric perspectives. The results evidence the successful integration of the image registration and RRN preprocessing steps by achieving a reasonable and stable performance.

Relaxation-Based Radiometric Normalization for Multitemporal Cross-Sensor Satellite Images

“…The histograms of the IR-MAD and proposed relaxation algorithm showed similar distributions to the original images, indicating that both normalization methods preserved the original distribution of the surface-reflectance values in the images. This suggests that both methods can successfully normalize images by maintaining the features and details in images during the normalization process [54].…”

LIRRN: Location-Independent Relative Radiometric Normalization of Bitemporal Remote-Sensing Images

“…To overcome the aforementioned limitation, keypoint-based RRN methods [19,20] have been developed that are robust to variations in scale, illumination, and viewpoints between subject and reference images. These methods typically operate in two steps.…”

Radiometric Normalization Using a Pseudo−Invariant Polygon Features−Based Algorithm with Contemporaneous Sentinel−2A and Landsat−8 OLI Imagery

“…Manual extraction allows pairwise pixels in different ground types to be selected, decreasing the matching error for paired pixels [7,15]. Although spatial mismatch problems in RRN are well−addressed in [18,33,34], three issues remain: (1) the spectra of paired pixels may differ due to the pixel−level spatial mismatch, especially for pairwise images with different sensing angles that create spatial distortion; (2) a regression analysis in which one variable does not consider the influence of adjacent bands, and; (3) applications are not focused on crop−growing areas and extracted PIFs are untrained to different types of crop or ground materials. To address these issues, we propose new relative radiometric normalization methods by replacing invariant points or pixels with invariant polygons or surfaces to apply to a crop−growing area that has been surveyed on the ground.…”