SearcharxivSearch

arXiv subjects

K. Gwinner

Publications and source records attributed to K. Gwinner.

2 recordsLinked to original sources

A global colour mosaic of Mars from Mars Express HRSC high altitude observations

The ever-changing transparency of the Martian atmosphere hinders the determination of absolute surface colour from spacecraft images. While individual high-resolution images from low orbit reveal numerous colour details of the geology, the colour variation between images caused by scattering off atmospheric dust can easily be of greater magnitude. The construction of contiguous large-scale mosaics has thus required a strategy to suppress the influence of scattering, often a form of high-pass filtering, which limits their ability to convey colour variation information over distances greater than the dimensions of single images. Here we use a dedicated high altitude observation campaign with the Mars Express High Resolution Stereo Camera (HRSC) (Neukum and Jaumann, 2004; Jaumann et al., 2007), applying a novel iterative method to construct a globally self-consistent colour model. We apply the model to colour-reference a high-altitude mosaic incor-porating long-range colour variation information. Using only the relative colour information internal to individual images, the influence of absolute image to image colour changes caused by scattering is minimised, while the model enables col-our variations across image boundaries to be self-consistently reconstructed. The resulting mosaic shows a level of colour detail comparable to single images, while maintaining continuity of colour features over much greater distances, thereby increasing the utility of HRSC colour images in the tracing and analysis of martian surface structures.

astro-ph.EP

Automated detection of block falls in the north polar region of Mars

We developed a change detection method for the identification of ice block falls using NASA's HiRISE images of the north polar scarps on Mars. Our method is based on a Support Vector Machine (SVM), trained using Histograms of Oriented Gradients (HOG), and on blob detection. The SVM detects potential new blocks between a set of images; the blob detection, then, confirms the identification of a block inside the area indicated by the SVM and derives the shape of the block. The results from the automatic analysis were compared with block statistics from visual inspection. We tested our method in 6 areas consisting of 1000x1000 pixels, where several hundreds of blocks were identified. The results for the given test areas produced a true positive rate of ~75% for blocks with sizes larger than 0.7 m (i.e., approx. 3 times the available ground pixel size) and a false discovery rate of ~8.5%. Using blob detection we also recover the size of each block within 3 pixels of their actual size.

cs.CV