2016/07/31 by J. Casado, Y. Ascasíbar, Y. Ascasibar +8 · 1 citation
Biochemistry, Genetics and Molecular Biology · Engineering · Environmental Science · Mathematics · Physics and Astronomy · #Algorithm #Artificial intelligence #Bayesian probability #CCD and CMOS Imaging Sensors #Computer science #Data set #Image (mathematics) #Mathematics #Noise (video) #Pattern recognition (psychology) #Physics #Remote Sensing in Agriculture #SIGNAL (programming language) #Segmentation #Set (abstract data type) #Spatial analysis #Spectroscopy Techniques in Biomedical and Chemical Research #Statistics #astro-ph.GA #astro-ph.IM
paper · pdf · doi:10.1093/mnras/stw3362
21 pages, 7 Figure, 4 Tables
arxiv created 2016/12/19 · openalex publication_date 2016/12/24 · arxiv updated 2017/01/18 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
This paper describes the Bayesian Technique for Multi-image Analysis (BATMAN), a novel image-segmentation technique based on Bayesian statistics that characterizes any astronomical data set containing spatial information and performs a tessellation based on the measurements and errors provided as input. The algorithm iteratively merges spatial elements as long as they are statistically consistent with carrying the same information (i.e. identical signal within the errors). We illustrate its operation and performance with a set of test cases including both synthetic and real integral-field spectroscopic data. The output segmentations adapt to the underlying spatial structure, regardless of its morphology and/or the statistical properties of the noise. The quality of the recovered signal represents an improvement with respect to the input, especially in regions with low signal-to-noise ratio. However, the algorithm may be sensitive to small-scale random fluctuations, and its performance in presence of spatial gradients is limited. Due to these effects, errors may be underestimated by as much as a factor of 2. Our analysis reveals that the algorithm prioritizes conservation of all the statistically significant information over noise reduction, and that the precise choice of the input data has a crucial impact on the results. Hence, the philosophy of BaTMAn is not to be used as a ‘black box’ to improve the signal-to-noise ratio, but as a new approach to characterize spatially resolved data prior to its analysis. The source code is publicly available at http://astro.ft.uam.es/SELGIFS/BaTMAn.