Automated detection of block falls in the north polar region of Mars. (January 2020)

• Record Type: Journal Article
• Authors: Fanara, L.; Gwinner, K.; Hauber, E.; Oberst, J.
• Is Part Of: Planetary and space science. Volume 180(2020)
• Publication Date: 2020-01
• Subjects: Change detection -- Block falls -- Machine learning -- Blob detection -- Mars; Space sciences -- Periodicals; Atmosphere, Upper -- Periodicals; Sciences spatiales -- Périodiques; Haute atmosphère -- Périodiques; 523
• DOI: 10.1016/j.pss.2019.104733
• Languages: English
• ISSNs: 0032-0633

Abstract

Abstract: 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 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 each consisting of 1000 × 1000 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 0.5 m 2 (i.e., approx. 3 times the available ground pixel size) and a false discovery rate of ~8.5%. Using blob detection, we were also able to recover the size of each block within 3 pixels of their actual size. Highlights: A change detection method for block falls' identification on Mars is proposed. The images and a training data set are prepared. A combination of SVM/HOG and blob detection is used. The results show a true positive rate of ~75% and a false detection rate of ~8.5%.