Evaluation of Satellite Image Correction Methods Caused by Differential Terrain Illumination

Abstract

The problem due to differential terrain illumination on satellite imagery is experienced by most of areas which are on mountainous terrain. This may cause variations in reflectance of similar ground features which lead to a misclassification of land cover classes due to different topographic positions. This phenomenon most commonly occurred in the areas which are located on southern and northern hemisphere because of the low sun inclination. This problem has been a major interest for researchers to be solved prior to the land cover classification process. For satellite images which experience this kind of problem, topographic correction need to be applied in order to reduce the illumination effects prior to land cover classification process. This research is aimed at conducting topographic correction of multi spectral SPOT satellite data as well as evaluating the three topographic correction methods. They are Cosine which is based on Lambertian reflectance assumption, as well as Minnaert correction and C correction methods which are based on non-Lambertian reflectance assumption. The data used in this study are two scenes of SPOT images of forested mountainous area of Miyazaki Prefecture, Kyushu, Japan. Research steps had been conducted in this study including geometric correction, sample data collection for calculating Minnaert constants and C constants at location which represents the whole study area, topographic correction for two scenes SPOT images, and results analysis. The results show that Cosine method did not show good performance for the study area which is topographically dominated by rugged terrain. Whereas Minnaert method and C method gave satisfactory results as is indicated by the statistical data as well as visual interpretation. However the Minnaert correction method showed slightly better performance than the C correction method.

Uji Performa Citra Worldview 3 dan Sentinel 2A untuk Pemetaan Kedalaman Laut Dangkal (Studi Kasus di Kepulauan Karimunjawa, Jawa Tengah)

“…Tidak hanya itu, citra dengan resolusi sedang open source (gratis) juga banyak tersedia salah satunya citra Sentinel 2A dengan resolusi spasial 10 m (Satellite Imaging Corporation, 2017). Kedua tipe ini memiliki kelebihan diantaranya revisit time yang cukup tinggi dan kanal spektral yang cukup lengkap yang dapat digunakan dalam berbagai keperluan seperti analisis biomassa dan pemetaan sumberdaya (Santosa, 2016;Bakara, 2014).…”

Klasifikasi Digital Tutupan Lahan Berbasis Objek menggunakan Integrasi Data Lidar dan Citra Satelit di Kawasan Tamalanrea Indah, Kota Makassar

“…Teknologi satelit optis memanfaatkan pencitraan menggunakan sensor pasif, yang memanfaatkan pantulan sinar matahari dari permukaan bumi ke sensor optik satelit sehingga memiliki kapabilitas yang berbeda dengan LiDAR. Jika wahana sensor aktif (radar/LiDAR) memiliki keunggulan dalam segi spasial (ketelitian geometris lebih tinggi tanpa adanya pergeseran relief), satelit memiliki keunggulan dalam segi spektral (rentang kanal spectral/band lebih luas dalam mengidentifikasi karakteristik pada objek) (Kushardono, 1997;Santosa, 2016;Aristalindra dkk, 2020).…”

Seasonal Waterlogging Variation in the Sylhet City Corporation Area, Bangladesh- an Integrated Geospatial Approach

“…The land/water edge is often perplexing in a single near-infrared (NIR) band, and therefore, two-band instead of single-band data such as green and NIR bands can be utilized in such cases. Thus, rationing of the two band data carries the advantage of the contrast in the reflectance of separate wavelengths in improving a particular feature from the multispectral satellite data (Chowdary et al, 2008;Santosa, 2016).…”