Multi-criteria evaluation of topographic correction methods

• 10 topographic correction (TOC) methods were assessed in 3 images of a study site.
• 7 evaluation strategies were compared and a multi-criteria evaluation was proposed.
• Statistic-Empirical method, C-Correction, and SCS + C performed the best.
• Real and synthetic images were used to assess the TOC methods.
• Multi-criteria analysis to assess the performance of TOC methods is recommended.

In the last decades, several topographic correction methods (TOC) have been proposed, but there is not an agreement on the best method. Furthermore, different evaluation criteria have been used in the past, and there is not any simple and objective evaluation procedure to measure the quality of the correction. Consequently, a multi-criteria analysis of widely used topographic correction methods is required that evaluates their performance over different sensors, terrain and temporal configurations. In this work, ten TOC methods were assessed using seven different evaluation strategies. The analysis was carried out for three SPOT5 images acquired over a mountainous area of northern Spain. The images had different acquisition dates and solar angles, so as to evaluate performance under varying illumination conditions. The results obtained showed that Statistic-Empirical method, C-Correction and Sun-Canopy-Sensor + C performed the best, and differences were minor when favorable illumination conditions were considered. For the seven tested evaluation strategies, interquartile range reduction of land covers or the comparison of sunlit and shaded slopes gave very similar results, whereas there were greater contrasts among other criteria.

Figure optionsDownload high-quality image (705 K)Download as PowerPoint slide